No description available.

To ensure safety regarding dust explosion hazards, it is important to study the dust lifting process experimentally and identify important parameters that will be valuable for development and validation of numerical predictions of this phenomenon. A new shock tube test section was developed and integrated into an existing shock tube facility. The test section allows for shadowgraph or laser scattering techniques to track dust layer particle motion. The test section is designed to handle an initial pressure of 1 atm with an incident shock wave velocity up to Mach 2 to mimic real world conditions. The test section features an easily removable dust pan and inserts to allow for adjustment of dust layer thickness. The design allows for the changing of experimental variables including initial pressure, Mach number, dust layer thickness and characteristics of the dust itself. A separate vacuum manifold was designed to protect existing equipment from negative side effects of the dust. A study was performed to demonstrate the capabilities of the new facility and to compare results with experimental trends formerly established in the literature. Forty-micron limestone dust with a layer thickness of 3.2 mm was subjected to Mach 1.22 and 1.38 shock waves, and a high-speed shadowgraph was used for flow visualization. Dust layer rise height was graphed with respect to shock wave propagation. Dust particles subjected to a Mach 1.38 shock wave rose more rapidly and to a greater height with respect to shock wave propagation than particles subjected to a Mach 1.22 shock wave. These results are in agreement with trends found in the literature, and a new area of investigation was identified.

The use of chemiluminescence measurements to monitor a range of combustion processes has been a popular area of study due to their reliable and cost-effective nature. Electronically excited carbon dioxide (CO2*) is known for its broadband emission, and its detection can lead to valuable information; however, due to its broadband characteristics, CO2* is difficult to isolate experimentally, and the chemical kinetics of this species is not well known. Although numerous works have monitored CO2* chemiluminescence, a full kinetic scheme for the species has yet to be developed. A series of shock-tube experiments was performed in H2-N2O-CO mixtures highly diluted in argon at conditions where emission from CO2* could be isolated and monitored. These results were used to evaluate the kinetics of CO2*, in particular, the main CO2* formation reaction, CO + O + M CO2* + M (R1). Based on collision theory, the quenching chemistry of CO2* was determined for eleven common collision partners. The final mechanism developed for CO2* consisted of 14 reactions and 13 species. The rate for R1 was determined based on low-pressure experiments performed in two different H2-N2O-CO-Ar mixtures. Final mechanism predictions were compared with the experimental results at low and high pressures, with good agreement seen at both conditions. Peak CO2* trends with temperature as well as overall CO2* species time histories were both monitored. Comparisons were also made with previous experiments in methane-oxygen mixtures, where there was slight over-prediction of CO2* experimental trends by the mechanism.Experimental results and mechanism predictions were also compared with past literature rates for CO2*, with good agreement for peak CO2* trends, and slight discrepancies in overall CO2* species time histories. Overall, the ability of the CO2* mechanism developed in this work to reproduce a range of experimental trends represents an improvement over existing models.

High-speed particle dispersion research is motivated by the energy release enhancement of explosives containing solid particles. In the initial explosive dispersal, a dense gas-solid flow can exist where the physics of phase interactions are not well understood. A dense particle flow is generated by the interaction of a shock wave with an initially stationary packed granular bed. The initial packed granular bed is produced by compressing loose aluminum oxide powder into a 6.35 mm thick wafer with a particle volume fraction of 0.48. The wafer is positioned inside the shock tube, uniformly filling the entire cross-section. This results in a clean experiment where no flow obstructing support structures are present. Through high-speed shadowgraph imaging and pressure measurements along the length of the channel, detailed information about the particle-shock interaction was obtained. Due to the limited strength of the Mach 2 incident shock wave, no transmitted shock wave is produced. The initial “solid-like” response of the particle wafer acceleration forms a series of compression waves that coalesce to form a shock wave. Breakup is initiated along the periphery of the wafer as the result of shear that forms due to the fixed boundary condition. Particle break-up starts at local failure sites that result in the formation of particle jets that extend ahead of the accelerating, largely intact, wafer core. In a circular tube the failure sites are uniformly distributed along the wafer circumference. In a square channel, the failure sites, and the subsequent particle jets, initially form at the corners due to the enhanced shear. The wafer breakup subsequently spreads to the edges forming a highly non-uniform particle cloud. / Thesis (Master, Mechanical and Materials Engineering) -- Queen's University, 2012-09-25 14:15:37.615

No description available.

No description available.

Early identification of mutant embryos prior to the manifestation of a defect facilitates the study of dysmorphogenesis. The In(l)lRk inversion was used as a cytogenetic marker to distinguish embryonic day 9 (D9) splotch (Sp) and splotch-delayed $(Sp sp{d})$ mouse mutants from heterozygous and wild-type littermates, and cellular aspects of abnormal neurulation and NCC migration were examined before inherent neural tube defects (NTDs) and deficiencies in neural crest cell (NCC) derivatives developed. In vitro analysis of NCC emigration from D9 neural tube explants revealed a delay in the release of NCCs from mutant neural tubes compared to controls, suggesting that the primary effect of the mutation was intrinsic to the neuroepithelium. Immunofluorescent localization of the neural cell adhesion molecule (N-CAM) antibody in situ demonstrated an increased intensity of antibody fluorescence in mutant tissue compared to controls, and further characterization by immunoblot analysis showed an altered embryonic N-CAM profile in both Sp and $Sp sp{d}$ mutants at D9 of gestation. The importance of N-CAMs in mediating cellular organization and communication has been well documented, supporting the idea that an alteration in this adhesion mechanism could result in the types of defects seen in splotch locus mouse mutants.

Cell adhesion molecules are known to play crucial roles in a variety of developmental processes. The neural cell adhesion molecule N-CAM is strongly implicated in neurulation and neural crest cell (NCC) migration and was thus studied in splotch (Sp) neural tube defect mutant embryos. At the 20 somite-stage of gestation day 9, Sp N-CAM was found to contain polysialic acid (PSA) side chains which are normally only present beginning at gestational day 11. Younger embryos at 12 and 14 somites also showed the presence of PSA on N-CAM, which was absent in controls. Enzymatic removal of PSA from N-CAM resulted in isoforms which migrated identically to PSA-free N-CAM isoforms in SDS-polyacrylamide gels. The post-translational modification of N-CAM appears to be the primary target of the Sp gene. In view of N-CAM's importance during development, an alteration at a critical stage is likely to result in the cascade of abnormalities seen in Sp mutants. / A new genotyping assay was also implemented for examination of N-CAM in Sp and other related wildtype strains.

Laminar, two-dimensional, constant-property numerical simulations of flat tube heat exchanger devices operating in flow regimes in which self-sustained oscillations occur were performed. The unsteady flow regimes were transition flow regimes characterized by cyclic variations of flow parameters such as stream-wise or cross-stream velocity. A computer code was developed to perform the numerical simulations. Spatial discretization was based upon a Control Volume Finite Element Method (CVFEM). Temporal discretization was based upon a semi-implicit Runge-Kutta method. Double Cyclic conditions were used to limit the numerical domains to one repeating geometric module. Nine geometric domains representing flat tube heat exchanger devices were tested over a range of Reynolds numbers. A maximum Reynolds number (Re) of 2000 was established to keep the study within the transition range. For each domain, a critical Reynolds number (Re_crit) was found such that for Re < Re_crit the flow was steady, laminar flow and for Re > Re_crit the flow exhibited cyclic oscillations. For the cases tested, the variation in longitudinal pitch had little impact on the Re_crit value for a fixed transverse pitch. However, for a fixed longitudinal pitch, the Re_crit was increased for decreasing transverse pitch. The results demonstrate the importance of using unsteady simulation methods for these cases. Nusselt numbers predicted by the unsteady method were on the order of 65% higher than predicted by steady methods for the same Reynolds numbers. Data for required pumping power versus resultant Nusselt number were collected which showed four distinct operating regions for these devices spanning the low Reynolds number, steady flow region through the self-sustained oscillating flow region. Based on the data, the recommended operating region is the region of self-sustained oscillations as this region is characterized by the highest increase in Nusselt number per increase in required pumping power.

In methanol processing plants, steam-methane reformers consist of hundreds of vertical tubes operating at temperatures up to 1000°C. These reformer tubes fail by creep through the formation of creep voids during service. Preliminary research showed that the occurrence of these voids was not random and may be related to certain microstructural features of the material. In the present research, the technique of serial sectioning was used to generate threedimensional reconstructions of voids in several steam-methane reformer tube samples with creep damage. The serial sectioning method and subsequent 3D reconstruction revealed creep void information such as size, density, location, and shape in three-dimensions, information that cannot be obtained from two-dimensional micrographs alone. Samples were obtained at various locations along the length of an ex-service reformer tube to investigate the effects of service conditions on the characteristics of creep voids. In addition, samples were taken from various positions along the wall thickness where there were differences in temperature, stress, and microstructure. Additionally, the identity and crystallographic orientations of the phases adjacent to creep voids were studied by electron backscatter diffraction (EBSD) to determine the crystallographic trends in creep void locations. Three dimensional observations revealed that creep voids were indeed not uniformly distributed through the volume in terms of their size, shape, and location. All voids appeared next to carbides and these voids came into contact with M₂₃C₆ precipitates somewhere along their perimeter. Most of the voids were found on austenite (ɣ) grain facets (the interface between two ɣ grains) but the larger voids were generally found at grain edges and corners. The grain boundaries where voids were located were generally oriented at 45 degrees with respect to the hoop stress direction. Here, the effective stress due to a combination of loading and temperature were highest. xviii Abstract EBSD results showed that 80% of the M₂₃C₆ precipitates surrounding these voids have an irrational crystallographic orientation relationship (OR) with the austenite matrix. In contrast, grain boundary precipitates in an aged sample always show a rational OR with respect to one adjacent grain. This implied that the preferred sites for creep voids are low registry boundaries between M₂₃C₆ precipitates and austenite. The data obtained from 3D observations were applied to a classic void growth model. Various permutations of the parameters obtained from this work were applied to the model to simulate conditions that may be beneficial to extending the service lives of reformer tubes. It was shown that the void growth model required accurate and representative materials constants for good estimation of life. Furthermore, the model revealed that more work was required in terms of observations of void nucleation in 3D, in order to fully utilize the model. Finally, it was shown that void density measurements are the most critical item for accurate prediction of growth of voids.