This thesis describes the development and validation of a modified eddy viscosity model to take into account the misalignment between stress a_{ij} and strain S_{ij} fields for reacting flow. The stress-strain misalignment is quantified by introducing a C_{as}=-a_{ij}S_{ij} /\sqrt{2S_{ij}S_{ij}} parameter. A new transport equation for C_{as} was derived from a full Reynolds stress model (RSM). The C_{as} transport equation was coupled to a standard EVM model (e.g. k-\omega SST) to form three equations model. This model is a new version of the SST-C_{as} model introduced by Revell (Revell2006), to incorporate buoyancy and combustion effects for buoyant reacting flow (e.g. fire). The performance of the proposed model was initially investigated via non-reacting buoyant plumes with different level of unsteadiness. The buoyant plumes were also simulated using different turbulence models and the results were compared to proposed model and experimental data. The model shows significant improvements for velocity and scalar profiles in region closed to plume centreline compared to the original SST model. The SST-C_{as} model was then applied for a real fire test case (Steckler room), and the results were compared to experimental data and results of RSM models. The SST-C_{as} model generally yields better than classical EVM models and reduces the gap between the RSM and EVM prediction with 25-30\% additional computational expenses. This work is still under development and validation for reacting flows, further work is going on to include the turbulence combustion interaction and validate it with DNS data.

Thesis (Ph. D.)--Joint Program in Oceanography (Massachusetts Institute of Technology, Dept. of Earth, Atmospheric, and Planetary Sciences; and the Woods Hole Oceanographic Institution), 2001. / "September 2001." Vita. Page 223 blank. / Includes bibliographical references. / This thesis studies interactions between mid-ocean ridges and mantle plumes using geophysics, geochemistry, and geodynamical modeling. Chapter 1 investigates the effects of the Marion and Bouvet hotspots on the ultra-slow spreading, highly-segmented Southwest Indian Ridge (SWIR). Gravity data indicate that both Marion and Bouvet impart high-amplitude mantle Bouguer anomaly lows to the ridge axis, and suggest that long-offset transforms may diminish along-axis plume flow. Building upon this observation, Chapter 2 presents a series of 3D numerical models designed to quantify the sensitivity of along-axis plume-driven mantle flow to transform offset length, spreading rate, and mantle viscosity structure. The calculations illustrate that long-offset transforms in ultra-slow spreading environments may significantly curtail plume dispersion. Chapter 3 investigates helium isotope systematics along the western SWIR as well as near a global array of hotspots. The first part of this study reports uniformly low 3He/4He ratios of 6.3-7.3 R/Ra along the SWIR from 9⁰-24⁰E, compared to values of 8 +/- 1 Ra for normal mid-ocean ridge basalt. The favored explanation for these low values is addition of (U+Th) into the mantle source by crustal and/or lithospheric recycling. Although high He/4He values have been observed along the SWIR near Bouvet Island to the west, there is no evidence for elevated 3He/4He ratios along this section of the SWIR. The second part of Chapter 3 investigates the relationship between 3He/4He ratios and geophysical indicators of plume robustness for nine hotspots. / (cont.) A close correlation between a plume's flux and maximum 3He/4He ratio suggests a link between plume upwelling strength and origination in the deep, relatively undegassed mantle. Chapter 4 studies 3D mantle flow and temperature patterns beneath oceanic ridge-ridge-ridge triple junctions (TJs). In non-hotspot-affected TJs with geometry similar to the Rodrigues TJ, temperature and upwelling velocity along the slowest-spreading of the three ridges are predicted to increase within a few hundred kilometers of the TJ, to approach those of the fastest-spreading ridge. Along the slowest-spreading branch in hotspot-affected TJs such as the Azores, a strong component of along-axis flow directed away from the TJ is predicted to advect a hotspot thermal anomaly away from its deep-seated source. / by Jennifer E. Georgen. / Ph.D.

Mineral chemistry and petrography of basalts from the Kimama drill core recovered by Hotspot: Snake River Scientific Drilling Project, Idaho establish crystallization conditions of these lavas. Twenty-three basalt samples, from 20 individual lava flows were sampled from the upper 1000 m (of the 1912 m drilled) core drilled on the axis of the Snake River Plain, and represent approximately 3 m.y. of volcanism (rocks at the bottom of the hole are ~6 Ma). Rock from the upper 1000 m are typically fresh, while those lower in the core are more altered and are less likely to preserve fresh phenocrysts to analyze. Intratelluric phenocrysts (pre-eruption) are: olivine, plagioclase and Cr-spinel inclusions in olivine and plagioclase; groundmass phases (post-eruption) are: olivine, plagioclase, clinopyroxene, magnetite and ilmenite. Olivine core compositions range from Fo84-68, plagioclase cores range from An80-62, clinopyroxene ranges in composition from Wo47-34, En47-28, Fs30-15, spinel inclusions are Cr (up to 20 wt % Cr2O3) and Al-rich (up to 35 wt % Al2O3) and evolve to lower concentrations of Cr and Al and higher Fe and Ti, chromian titanomagnetite to magnetite, and ilmenite are groundmass oxide phases. Thermobarometry of Kimama core basalts indicates that the phenocryst phases crystallized at temperatures of 1155 to 1255°C at depths of 7 to 17 km, which is within or near the seismically imaged mid-crustal sill. Plagioclase hygrometry suggests that these lavas are relatively anhydrous with less than 0.4 wt % H2O. Groundmass phases crystallized at lower temperatures (<1140°C) after eruption. Oxygen fugacity inferred from Fe-Ti oxide equilibria is at or just below the QFM buffer. The origin of the basaltic rocks of the Snake River Plain has been attributed to a mantle plume or to other, shallow mantle processes. Mineral and whole rock major and trace element geochemistry of the olivine tholeiites from the Kimama core are used to distinguish between these two sources (deep or shallow mantle). Whole rock compositions were corrected for plagioclase and olivine fractionation to calculate primary liquids to estimate mantle potential temperatures. Olivine phenocrysts have the pyroxenite source characteristics of low Mn and Ca, but a peridotite source characteristic of low Ni. Thus, trace element models were used to test whether there is pyroxenite in the source of the Snake River Plain basalts, as hypothesized for Hawaii and other plume-related hotspots (e.g., Sobolev et al., 2005; Herzberg, 2011). Olivine chemistry and trace element models establish that the basalt source is a spinel peridotite, not a pyroxenite. The average mantle potential temperature obtained for these samples is 1577°C, 177°C hotter than ambient mantle, suggesting that the basaltic liquids were derived from a thermal plume. Silica activity barometry shows that melt segregation occurs between 80 and 110 km depth, which is within or very near the spinel stability field, and suggests that the lithosphere has been eroded by the plume to a maximum depth of 80 km, and recent mantle tomography suggests that it may be even thinner.

Volcanic eruptions pose hazards to life and insfrastructure, and contribute to the resurfacing of earth and other planetary bodies. Lavas and magmas are multi-phase suspensions of silicate melts (liquids), solid crystals, and vapor bubbles, and solidify into glass and rock upon cooling. The interactions between phases place important controls on the dynamics and timescales of magma and lava transport and emplacement. The purpose of this thesis is to explore the role of multiphase interactions in controlling eruption dynamics and inform conceptual and numerical models for hazard prediction. In Chapters 1 and 2, centimeter to meter scale analog experiments are used to explore the multi-phase rheological properties and flow behaviors of bubble- and particle-bearing suspensions. Optical imaging of dam-break experiments presented in Chapter 1 expand existing experimental parameter ranges for lava analogs to higher bubble concentrations than existing datasets (up to 82% by volume bubbles and 37% by volume particles). I develop a constitutive relationship for threephase relative viscosity, and demonstrate that at the low strain-rate conditions relevant to many natural lava flows, accounting for the rheological effect of bubbles can result in the prediction of slower runout speeds. Chapter 2 expands upon the work of Chapter 1 using different analog materials observed using nuclear magnetic resonance imaging (MRI) phase-contrast velocimetry (PCV) to measure velocity in the flow interior of three-phase dam-break experiments. I find that for high-aspect ratio particles (sesame seeds), phase segregation into shear bands readily occurs, even at low particle fraction (20%) and results in strain localization. I suggest that the presence of shear bands can lead to faster flow runout than predicted using assumptions of bulk rheology. Chapter 3 analyzes thermal infrared (IR) time-lapse photography and videography of Hawaiian to Strombolian explosive activity during the 2021 eruption of Cumbre Vieja volcano, La Palma, Canary Islands, Spain. Images are analyzed to find time series of apparent plume radius, velocity, and apparent volume flux of high-temperature gas and lava. I compare with other measures of eruptive activity, including remote observations of plume height, SO₂ flux, effusive flux, tremor, and events at the volcano edifice including edifice collapses and the opening of new vents. I find correlations between tremor and explosive flux, but no correlation with SO2 flux or effusive flux, which I interpret as evidence of bubble segregation, highlighting the role of phase segregation and temporal variability in material properties in natural systems. Finally, in Chapter 4, I develop a novel finite element model to explore the interaction between a viscous flow with a solidified crust, and the effect of these interactions on lava flow and lava dome emplacement. I develop a model that couples a temperature-dependent viscous interior with an elastic shell flowing into air, water, or dense atmospheres. The model expands upon existing numerical simulations used in volcanology to have direct applications to lava flows and domes on the sea floor, which accounts for a large portion of the volcanism on Earth, and volcanism on other planetary bodies. Additionally, the formation of levees or solidified flow fronts that fracture and lead to a restart of flow. These lava flow breakouts pose a significant hazard, but there are currently no volcanological community codes capable of using a physics-based approach to predict the timing or location of breakouts. The model in Chapter 4 is the first to allow for assessment of the likelihood of failure at the scale of a flow lobe. Chapter 4 describes the model formulation and verification, and validation against centimeter-scale molten basalt experiments. The dissertation as a whole integrates work using a variety of methods including analog experiments, observations of natural eruptions, and numerical simulations to contribute to our understanding of the effects of multi-phase interactions on volcanic eruptions.

No description available.

In recent years, as control systems for boiler emissions have been upgraded, some utility sources have experienced increased plume opacity. Cases of plume opacity exceeding in-stack opacity are due to 1) the aerosol formed by condensation of primary sulfuric acid and water vapor onto polydisperse plume particles and 2) the presence of fine particles which grow into the visual size range by heterogeneous condensation and coagulation processes as the plume is cooled and diluted by mixing with the ambient air. In order to better understand the factors leading up to acid plume formation, a computer simulation model has been developed. This plume opacity model has been utilized to simulate sulfuric acid aerosol formation and growth. These processes result from homogeneous nucleation, condensation and coagulation which substantially increase the concentration of submicrometer sized aerosols. These phenomena bring about significant increases in plume opacity. Theoretical relationships have been derived and transformed into 21 computer model to predict plume opacity at various downwind distances resulting from pulverized coal combustion operations. This model consists of relatively independent components-such as an optics module, a bimodal particle size distribution module, a polydisperse coagulation module, a vapor condensation and nucleation module and a plume dispersion module-which are linked together to relate specific flue gas emissions and meterological conditions to plume opacity. This unique, near-stack, plume-opacity-model approach provides an excellent tool for understanding and dealing with such complex issues as: • increasing plume opacity observed for emissions containing sulfuric acid aerosols, • explaining the correlation between primary particle size distribution and light—scattering effects, • predicting the opacity level resulting from combustion of various coal types, • predicting control equipment effects on plume opacity. / Ph. D.

Among the most ubiquitous and persistent air quality problems facing urban areas are high concentrations of gas phase oxidants and fine particulate matter. Ozone and particulate matter concentrations in urban areas are significantly influenced by other factors in addition to local emissions, such as regional transport spanning distances as large as 1000 kilometers. Despite the importance of regional transport in meeting air quality standards, to date most analyses of regional transport have focused only on short duration episodes, or semi-quantitative assessments. The development and evaluation of seasonal, quantitative assessments of regional pollutant transport, based on modeling calculations and observational data is the topic of this dissertation. The observational data available through the Texas Air Quality Studies in 2000 and 2006 provide a unique opportunity to develop, evaluate, and improve methods for characterizing regional air pollutant transport. Measurements collected during these studies are used as the primary observational basis for characterizing regional ozone transport and to evaluate the performance of photochemical models. Results suggest that measurements (from aircraft and surface monitors) and the photochemical model provide consistent estimates of the magnitude of ozone transport. On this basis, photochemical modeling is used to determine potential impacts of regional ozone transport in Texas, under varying meteorological and photochemical conditions, as well as to characterize the dominant chemical and physical processes within urban plumes. While qualitative studies and limited quantitative analyses have been performed to assess regional ozone transport, this work includes the first detailed quantitative characterization of the importance of ozone transport over the course of an entire ozone season using both photochemical modeling and ambient data. Results demonstrate that urban plumes in Texas are capable of transporting significant amounts of ozone over distances spanning hundreds of kilometers. Furthermore, on a seasonal basis, there are a number of days characterized by high contributions from inter-city transport coinciding with high total ozone concentrations, suggesting that the role of inter-city transport will remain significant for many areas to demonstrate attainment of the NAAQS for ozone. Results also indicate that reductions in the impacts of inter-city transport are possible by decreases in emissions from source regions. / text

No description available.

Doctor of Philosophy / Department of Mechanical and Nuclear Engineering / M. H. Hosni / B. W. Jones / This dissertation focuses on airflow distribution in the longitudinal direction of a wide-body mockup aircraft cabin, turbulence energy and dissipation rates, and the effect of thermal plumes, generated by passengers, on airflow distribution within the cabin. The mockup cabin utilized for this study mimics a Boeing 767 passenger cabin and includes 11 rows in the longitudinal direction with each row consisting of seven seats. Each seat is occupied by an inflatable manikin which is instrumented with a 10 meters long wire heater generating approximately 100 Watts of distributed sensible heat, representing heat load from a sedentary human being. In order to investigate the fluid dynamics characteristics of the airflow within the cabin, different experimental techniques were implemented. Smoke visualization was used to qualitatively visualize the general airflow pattern inside the cabin. A tracer gas composed mainly of carbon dioxide was used to track the airflow distribution inside the cabin. The tracer gas was released in several locations and then sampled at various locations throughout the mockup cabin. The release and sampling of the tracer gas allowed tracing the airflow inside the cabin using non dispersive infrared sensors. Combining results from different release-sampling scenarios gave better understanding of the chaotic and three-dimensional nature of the airflow behavior inside the cabin. Air speed and turbulence parameters were evaluated using omni-directional probes. Finally, the effect of the heat generated by the thermal manikins on the airflow behavior was investigated. The results from the airflow visualization and the tracer gas were complementary and showed that there were multiple air circulations along the length of the cabin. The dimension of the circulations were controlled by the minimum physical distance inside the cabin. The identified-isotropic turbulence were spread over the full width of the cabin in the front and middle sections of the cabin, whereas, multiple-smaller circulations were identified in the rear section. Cabin sections identified with high speed fluctuations were associated with higher turbulence kinetic energy levels and lower local dissipation rates. These sections served as driving forces to create the circulations identified in the tracer gas experiments. Furthermore, the heat generated by the thermal manikins was shown to significantly impact the behavior of the gaseous flow inside the cabin, the turbulence parameters, and speed fluctuations. Detailed uncertainty analysis was conducted to estimate the uncertainty limits for the measurements taken. The uncertainty estimates obtained for the tracer gas results ranged from ±14% for the test cases with the heated manikins to ±17% with the corresponding unheated manikins cases. The data uncertainty limits for the turbulence parameters were of higher levels due to limitations associated with the omni-directional probes used to measure the speed. With flow repeatability phenomena in same locations inside the mockup cabin during different days reaching up to ±10%, the uncertainty estimates were considered acceptable for these chaotic and highly random airflow conditions within the cabin.

The rectangular pulsed plasma thruster (PPT) is an electromagnetic thruster that ablates Teflon propellant to produce thrust in a discharge that lasts 5-20 microseconds. In order to integrate PPTs onto spacecraft, it is necessary to investigate possible thruster plume-spacecraft interactions. The PPT plume consists of neutral and charged particles from the ablation of the Teflon fuel bar as well as electrode materials. In this thesis a novel application of quadruple Langmuir probes is implemented in the PPT plume to obtain electron temperature, electron density, and ion speed ratio measurements (ion speed divided by most probable thermal speed). The pulsed plasma thruster used is a NASA Glenn laboratory model based on the LES 8/9 series of PPTs, and is similar in design to the Earth Observing-1 satellite PPT. At the 20 J discharge energy level, the thruster ablates 26.6 mg of Teflon, creating an impulse bit of 256 mN-s with a specific impulse of 986 s. The quadruple probes were operated in the so-called current mode, eliminating the need to make voltage measurements. The current collection to the parallel to the flow electrodes is based on Laframboise's theory for probe to Debye length ratios between 5 and 100, and on the thin-sheath theory for ratios above 100. The ion current to the perpendicular probe is based on a model by Kanal and is a function of the ion speed ratio, the applied non-dimensional potential and the collection area. A formal error analysis is performed using the complete set of nonlinear current collection equations. The quadruple Langmuir probes were mounted on a computer controlled motion system that allowed movement in the radial direction, and the thruster was mounted on a motion system that allowed angular variation. Measurements were taken at 10, 15 and 20 cm form the Teflon fuel bar face, at angles up to 40 degrees off of the centerline axis at discharge energy levels of 5, 20, and 40 J. All data points are based on an average of four PPT pulses. Data analysis shows the temporal and spatial variation in the plume. Electron temperatures show two peaks during the length of the pulse, a trend most evident during the 20 J and 40 J discharge energies at 10 cm from the surface of the Teflon fuel bar. The electron temperatures after the initial high temperature peak are below 2 eV. Electron densities are highest near the thruster exit plane. At 10 cm from the Teflon surface, maximum electron densities are 1.04e20 ± 2.8e19 m-3, 9.8e20 ± 2.3e20 m-3, and 1.38e21 ± 4.05e20 m-3 for the 5 J, 20 J and 40 J discharge energy, respectively. The electrons densities decrease to 2.8x1019 ± 8.9e18 m-3, 1.2e20 ± 4.2e19 m-3, and 4.5e20 ± 1.2e20 m-3 at 20 cm for the 5 J, 20 J, and 40 J cases, respectively. Electron temperature and density decrease with increasing angle away from the centerline, and with increasing downstream distance. The plume is more symmetric in the parallel plane than in the perpendicular plane. Ion speed ratios are lowest near the thruster exit, increase with increasing downstream distance, but do not show any consistent angular variation. Peak speed ratios at a radial distance of 10 cm are 5.9±3.6, 5.3±0.39, and 4.8±0.41 for the 5 J, 20 J and 40 J discharge energies, respectively. The ratios increase to 6.05±5.9, 7.5±1.6, and 6.09±0.72 at a radial distance of 20 cm. Estimates of ion velocities show peak values between 36 km/s to 40 km/s, 26 km/s to 30 km/s, and 26 km/s to 36 km/s for the % J, 20 J, and 40 J discharge energies, respectively.