This dissertation presents three observational projects designed to characterize the outer disks of nearby galaxies (beyond the optical radius R₂₅). Until very recently, outer disks remained an elusive and poorly-understood component of disk galaxies. We first present a Large Binocular Telescope (LBT) optical imaging survey of nearby outer disks to examine the basic properties of this component. Our LBT observations indicate that most nearby galaxies host an outer disk with star formation occurring at a very low level. We detect hundreds of outer disk star clusters and show that they typically have masses ∼ 10² − 10⁴M⊙ and ages up to a Gyr. The clusters are born in groups that can remain clustered for a Gyr or more, while the clusters slowly evaporate stars into a diffuse stellar component. The clusters appear to form from localized overdensities in the gas distribution primarily associated with spiral structure. The clusters extend to 2R₂₅ in our sample. We find that some clusters may also reside well outside of their host galaxy’s gas disk. Our second project is a kinematic study of Hɑ knots in the outer disk of the large, isolated, face-on galaxy NGC 628, using Inamori Magellan Areal Camera and Spectrograph (IMACS) observations from the Magellan telescope. This galaxy shows a kinematically cold outer disk (velocity dispersion < 11 km s⁻¹) with a mass density ∑ = 7.5 M⊙ pc⁻². Our observations cannot exclude uniform star formation lasting a Hubble time in this outer disk and confirm that this component is an extension of the kinematically-cold inner disk. Our third project is a search for molecular emission in the outer disk of NGC 628, using the sensitive Atacama LargeMillimeter Array (ALMA) receiver on the Submillimeter Telescope (SMT). We did not detect emission from our outer disk pointings, though we are able to provide useful estimates for future ALMA observations of outer disk knots. Our SMT observations indicate that the H₂ / H I ratio is ∼ 100× lower in the outer disk than in the inner disk, which likely explains, at least in part, the trend towards smaller clusters and lower star formation rates at larger radii.

This study develops the foundation of a digitally implemented control system for the radial positioning of the read/write heads of an optical disk drive system. Topics addressed are: sample rate selection, state reconstruction, closed-loop system response, disk track-following filters, and measurement noise filters. Consideration is given to the unmodeled dynamic's influence on system performance, system sensitivity to parameter variations, and a one sample computational delay. What has been concluded from this work is that the system must be further desensitized to parameter variations, and, at this stage of the development of the control scheme, neither a computational delay nor the unmodeled dynamics degrade system performance significantly.

No description available.

Master of Science / Department of Chemical Engineering / Larry E. Erickson / Caprolactam (C₆H₁₁NO) is produced in industry primarily as a monomer to be converted to nylon-6 via a polymerization reaction. More demanding purity requirements for nylon-6 have increased the performance requirements of extraction columns in the purification train of caprolactam production. Caprolactam is produced by performing a Beckmann Rearrangement on cyclohexanone oxime followed by a neutralization of the excess oleum post reaction. The resulting side product is ammonium sulfate in water with a residual amount of caprolactam that has to be extracted with benzene from the aqueous ammonium sulfate solution to reduce product losses. The aqueous caprolactam liquor is extracted in another column into benzene which purifies the product from water soluble impurities. The resulting caprolactam dissolved in benzene is back extracted into water where the final purification steps are completed prior to storage. Rotating disk contactor (RDC) extraction columns were invented by Royal Dutch Shell in the early 1950’s. The columns have a rotor in the center that is driven by an electric motor to rotate equally spaced flat disks inside the column. There are equally spaced annulus shaped stators that serve to provide mixing-separation compartments for each of the mounted disks on the rotor. Of the variables to consider for the optimum performance of the extraction in the RDC extraction column is the rotor speed. Rotor speed curves are generated for the 3 RDC extraction columns of the caprolactam purification as well as calculations of the number of theoretical stages for each of the columns based on actual performance data. Benzene is the solvent of choice in this purification process however recent push by environmental groups and agencies as well as tightening regulations have driven a desire to find a more benign alternative to benzene for this process. A review of the research and literature on potential alternative solvents for caprolactam purification is summarized with positives and drawbacks for each possible alternative.

The focus of this thesis is on systems that employ both flash and magnetic disks as storage media. Considering the widely disparate I/O costs of flash disks currently on the market, our approach is a cost-aware one: we explore techniques that exploit the I/O costs of the underlying storage devices to improve I/O performance. We also study the asymmetric I/O properties of magnetic and flash disks and propose algorithms that take advantage of this asymmetry. Our work is geared towards database systems; however, most of the ideas presented in this thesis can be generalised to any data-intensive application. For the case of low-end, inexpensive flash devices with large capacities, we propose using them at the same level of the memory hierarchy as magnetic disks. In such setups, we study the problem of data placement, that is, on which type of storage medium each data page should be stored. We present a family of online algorithms that can be used to dynamically decide the optimal placement of each page. Our algorithms adapt to changing workloads for maximum I/O efficiency. We found that substantial performance benefits can be gained with such a design, especially for queries touching large sets of pages with read-intensive workloads. Moving one level higher in the storage hierarchy, we study the problem of buffer allocation in databases that store data across multiple storage devices. We present our novel approach to per-device memory allocation, under which both the I/O costs of the storage devices and the cache behaviour of the data stored on each medium determine the size of the main memory buffers that will be allocated to each device. Towards informed decisions, we found that the ability to predict the cache behaviour of devices under various cache sizes is of paramount importance. In light of this, we study the problem of efficiently tracking the hit ratio curve for each device and introduce a lowoverhead technique that provides high accuracy. The price and performance characteristics of high-end flash disks make them perfectly suitable for use as caches between the main memory and the magnetic disk(s) of a storage system. In this context, we primarily focus on the problem of deciding which data should be placed in the flash cache of a system: how the data flows from one level of the memory hierarchy to the others is crucial for the performance of such a system. Considering such decisions, we found that the I/O costs of the flash cache play a major role. We also study several implementation issues such as the optimal size of flash pages and the properties of the page directory of a flash cache. Finally, we explore sorting in external memory using external merge-sort, as the latter employs access patterns that can take full advantage of the I/O characteristics of flash memory. We study the problem of sorting hierarchical data, as such is necessary for a wide variety of applications including archiving scientific data and dealing with large XML datasets. The proposed algorithm efficiently exploits the hierarchical structure in order to minimize the number of disk accesses and optimise the utilization of available memory. Our proposals are not specific to sorting over flash memory: the presented techniques are highly efficient over magnetic disks as well.

Mon travail de thèse a porté sur la simulation numérique du comportement dynamique de poussières et de petits corps autour d’étoiles, appliqué à deux grands domaines de localisation : les disques de débris et les exozodis.Concernant les disques de débris, je me suis plus particulièrement intéressé aux arches mouvantes observées sur une période de 15 ans au sein du disque d’AU Mic. En supposant qu’elles proviennent toutes d’un corps parent unique et que les grains composant les arches ont la même dynamique que les arches elles-mêmes, j’ai montré que le corps parent doit être plus interne que la ceinture de planétésimaux (<25 UA) et qu’il peut être sur une orbite circulaire ou fixe par rapport à l’observateur. Afin d’expliquer la vitesse apparente des arches, il est nécessaire que les grains les composant soient submicrométriques pour être suffisamment sensibles à la pression du vent stellaire. Le champ magnétique à grande échelle de cette étoile est suffisant pour expliquer l’élévation verticale des arches, mais l’interaction des grains avec ce champ nécessite des études plus approfondies.D’autre part, j’ai développé un code numérique permettant de tester l’origine dynamique des poussières constituant les exozodis, en comparant les résultats de mes simulations aux observations. J‘ai montré que le scénario classique de migration par PR-drag de grains provenant d’une ceinture externe froide produit trop de flux dans le moyen infrarouge en regard du proche infrarouge, et cet effet n’est pas suffisamment contrebalancé par l’accumulation proche de la distance de sublimation. En revanche, le scénario cométaire, avec un apport de matière au plus près de l’étoile, permet de modérer le flux en moyen infrarouge. Les observations peuvent être reproduites avec une dizaine de comètes kilométriques autour de chaque étoile. Le code que j’ai conçu est capable d’appréhender de nombreux effets physiques, et il est possible de tester l’influence du DDE, de la pression du vent stellaire ou encore du champ magnétique.Par mon travail, j’ai montré que la prise en compte de la dynamique des grains de poussière permet de contraindre les propriétés physiques des grains, et j’ai développé des outils numériques adaptables à de nombreux cas de figures afin de pouvoir caractériser la diversité et la complexité de la poussière observée autour des étoiles. / During my thesis, I numerically simulated the dynamical behaviour of dust and small grains around stars, applied to two ranges of stellar distance : debris disks and exozodis.Concerning debris disks, I focused on the fast moving arch-like structures observed over 15 years inside the disk of AU Mic. Supposing that they all come from a single parent body and that the dust composing the arches have the same dynamics, I showed that the parent body must be closer-in than the planetesimal belt (<25 au), on a circular orbit or static with respect to the observer. Grains must be submicronic in order to explain the apparent velocity of the arches, gained by means of stellar wind pressure. Large-scale magnetic field of the star is large enough to explain the vertical extent of the arches, but require further investigations.I also developped a numerical code in order to test the dynamical origin of dust composing exozodis, to compare its results to the observations. I showed that the classical PR-drag scenario involving grains drifting inward from a distant cold parent belt produce an excess flux in mid-infrared compared to the near-infrared, unbalanced by the pile-up. The cometary scenario, by producing dust very close to the star, emits less in mid-infrared. A ten of kilometric comets can reproduce the flux levels observed around all stars. This numerical code is also able to handle the DDE, the stellar wind pressure or the magnetic field.During my work, I showed that taking into account the dust dynamics can constrain the dust physical properties. I developped adaptative numerical tools that can handle the variety and complexity of dust observed around stars.

Colloidal suspensions of disk-like particles have been of interest in both colloidal and liquid crystal studies because they exhibit unique liquid crystalline phases different from those of rod-like molecules. Disk-like particles, such as asphaltenes in heavy oil industry, clay particles in agriculture, and red blood cells in biology, are of great interest in a variety of industries and scientific areas. However, to fabricate monodisperse microdisks, uniform in structure or composition with precise control of particle size and shape has not yet succeeded. In this thesis, we show an experimental strategy of using microfluidic technique to fabricate homogeneous α-eicosene microemulsions with chloroform in an aqueous solution of sodium dedecyl sulfate (SDS). The monodisperse chloroform emulsions, generated by the glass-based microfluidic devices, ensure the precise control on microdisk particle size and shape. A systematic investigation was performed to study the relation between the resulted microdisk size and the initial concentration of α-eicosene in chloroform before evaporation. The smectic liquid crystalline phase inside the wax particles controls the coin-like disk shape below the melting temperature of wax’s rotator phase. The kinetics of the disk formation is observed using a polarized light microscope. Dynamic light scattering is used to characterize the Brownian motion of the microdisks, and the rotational diffusion is estimated from the image sequences taken by the charge-coupled device (CCD) camera. Effort has been put into collecting a large quantity of microdisks to investigate the discotic liquid crystalline phases, which can be readily probed by light scattering and microscope. In comparison, X-ray and neutron have to be used for the atomic liquid crystalline phase investigation.

Higher data rate hard disk drives(HDD) and improved read channel electronics are demanding preamplifier performance be extended well beyond 1 Gb/s. Historically, preamplifier power requirements were of low priority; however, with increased demand for battery powered devices such as laptops, MP3 players, personal video recorders, andmanyother wireless hand-held devices, power consumption has become an important design parameter.Furthermore, in order to continue to increase drive capacities, new read-write head technologies(vertical recording and TGMR heads) are demanding innovative preamplifier circuitsolutions.Today's production preamplifiers possess a wide band response of 2.5 MHz-600 MHz; however next generation preamplifiers willrequire response greater than 250 KHz-1 GHz.Low corner frequencies below 250 KHz present read recovery (sleep-to-read, write-to-read, etc) challenges which can limit drive capacity. This project targets a > 2 Gb/s TGMR (tunneling giantmagneto-resistive) read path for verticalrecording HDDs. A high performance BiCMOS process (IBM's 0.5?m 5HP process)is essential due to the large transconductances, low noise and highspeed requirements of the read path's first stage. System frequency limitations at the input are a result of the large TGMR read sensor and preamplifier input capacitance. Due to read head and preamplifier manufacturingvariations, resistive feedbackaroundthe firststage is usedto seta controlled input impedance targeted to match the interconnect transmission line. Head resistance variations lead to gain variations; however, the TGMR element becomes more sensitive with larger resistance. This, to a first order approximation, acts like an automatic gain control and reduces variations in gain due to the head.

Recently, Handheld devices are more and more popular. Most of them aim at the low price, small size, high computing power, and powerful functionalities. Therefore, the need for embedded operating systems in the market is absolutely vital. There are many embedded systems in the market, but embedded Linux has some advantages to be outstanding and widely accepted. For example, it has no proprietary problem and high portability, and is comparatively easy to be reconstructed and to develop new applications. Best of them all, it is open source software. Embedded systems are usually diskless systems. In order to keep permanent data in embedded Linux, using flash disk as its disk system is a widely adapted strategy. We use MTD (Memory Technology Devices) system to emulate flash memory as flash disk mounted into Linux virtual file system. This allows accessing flash memory with standard I/O operations without any extra effort. MTD system contains, ¡§user¡¨ and driver¡¨, two different modules. In driver modules, we use CFI (Common Flash Interface) to probe the flash chip and then partition it, while we use MTD BLOCK to emulate the flash partitions as block devices in user modules, and then mount them into Linux¡¦s virtual file system with JFFS2 (Journaling Flash File System version 2) type, which is a file system type specifically designed for flash devices according to the features of flash devices. The purpose of this thesis is to use MTD system to emulate Am29LV320DB flash chip as a flash disk in embedded Linux running on an ARM-based developing board, SMDK2410, designed by Samsung. I hope this porting can help the development of other advanced applications and provide an empirical platform for the research of embedded systems.

Higher data rate hard disk drives(HDD) and improved read channel electronics are demanding preampli&#64257;er performance be extended well beyond 1 Gb/s. Historically, preampli&#64257;er power requirements were of low priority; however, with increased demand for battery powered devices such as laptops, MP3 players, personal video recorders, andmanyother wireless hand-held devices, power consumption has become an important design parameter.Furthermore, in order to continue to increase drive capacities, new read-write head technologies(vertical recording and TGMR heads) are demanding innovative preampli&#64257;er circuitsolutions.Today's production preampli&#64257;ers possess a wide band response of 2.5 MHz-600 MHz; however next generation preampli&#64257;ers willrequire response greater than 250 KHz-1 GHz.Low corner frequencies below 250 KHz present read recovery (sleep-to-read, write-to-read, etc) challenges which can limit drive capacity. This project targets a > 2 Gb/s TGMR (tunneling giantmagneto-resistive) read path for verticalrecording HDDs. A high performance BiCMOS process (IBM's 0.5?m 5HP process)is essential due to the large transconductances, low noise and highspeed requirements of the read path's &#64257;rst stage. System frequency limitations at the input are a result of the large TGMR read sensor and preampli&#64257;er input capacitance. Due to read head and preampli&#64257;er manufacturingvariations, resistive feedbackaroundthe &#64257;rststage is usedto seta controlled input impedance targeted to match the interconnect transmission line. Head resistance variations lead to gain variations; however, the TGMR element becomes more sensitive with larger resistance. This, to a &#64257;rst order approximation, acts like an automatic gain control and reduces variations in gain due to the head.