Multilevel Latent Class Models with Dirichlet Mixing Distribution

Latent class analysis (LCA) and latent class regression (LCR) are widely used for modeling multivariate categorical outcomes in social sciences and biomedical studies. Standard analyses assume data of different respondents to be mutually independent, excluding application of the methods to familial and other designs in which participants are clustered. In this paper, we develop multilevel latent class model, in which subpopulation mixing probabilities are treated as random effects that vary among clusters according to a common Dirichlet distribution. We apply the Expectation-Maximization (EM) algorithm for model fitting by maximum likelihood (ML). This approach works well, but is computationally intensive when either the number of classes or the cluster size is large. We propose a maximum pairwise likelihood (MPL) approach via a modified EM algorithm for this case. We also show that a simple latent class analysis, combined with robust standard errors, provides another consistent, robust, but less efficient inferential procedure. Simulation studies suggest that the three methods work well in finite samples, and that the MPL estimates often enjoy comparable precision as the ML estimates. We apply our methods to the analysis of comorbid symptoms in the Obsessive Compulsive Disorder study. Our models' random effects structure has more straightforward interpretation than those of competing methods, thus should usefully augment tools available for latent class analysis of multilevel data.

Nonstructural proteins NS2-3 and NS4A of classical swine fever virus: essential features for infectious particle formation

The nonstructural protein NS2-3 of pestiviruses undergoes tightly regulated processing. For bovine viral diarrhea virus it was shown that uncleaved NS2-3 is required for infectious particle formation while cleaved NS3 is essential for genome replication. To further investigate the functions of NS2-3 and NS4A in the pestivirus life cycle, we established T7 RNA polymerase-dependent trans-complementation for p7-NS2-3-4A of classical swine fever virus (CSFV). Expression of NS2-3 and NS4A in trans restored the production of infectious particles from genomes lacking NS2-3 expression. Co-expression of cleaved NS4A was essential. None of the enzymatic activities harbored by NS2-3 were required for infectious particle formation. Importantly, expression of uncleavable NS2-3 together with NS4A rescued infectious particles from a genome lacking NS2, demonstrating that cleaved NS2 per se has no additional essential function. These data indicate that NS2-3 and NS3, each in association with NS4A, have independent functions in the CSFV life cycle.

Diesel spray mixing limited vaporization with non-ideal and multi-component fuel thermophysical property effects

Internal combustion engines are, and will continue to be, a primary mode of power generation for ground transportation. Challenges exist in meeting fuel consumption regulations and emission standards while upholding performance, as fuel prices rise, and resource depletion and environmental impacts are of increasing concern. Diesel engines are advantageous due to their inherent efficiency advantage over spark ignition engines; however, their NOx and soot emissions can be difficult to control and reduce due to an inherent tradeoff. Diesel combustion is spray and mixing controlled providing an intrinsic link between spray and emissions, motivating detailed, fundamental studies on spray, vaporization, mixing, and combustion characteristics under engine relevant conditions. An optical combustion vessel facility has been developed at Michigan Technological University for these studies, with detailed tests and analysis being conducted. In this combustion vessel facility a preburn procedure for thermodynamic state generation is used, and validated using chemical kinetics modeling both for the MTU vessel, and institutions comprising the Engine Combustion Network international collaborative research initiative. It is shown that minor species produced are representative of modern diesel engines running exhaust gas recirculation and do not impact the autoignition of n-heptane. Diesel spray testing of a high-pressure (2000 bar) multi-hole injector is undertaken including non-vaporizing, vaporizing, and combusting tests, with sprays characterized using Mie back scatter imaging diagnostics. Liquid phase spray parameter trends agree with literature. Fluctuations in liquid length about a quasi-steady value are quantified, along with plume to plume variations. Hypotheses are developed for their causes including fuel pressure fluctuations, nozzle cavitation, internal injector flow and geometry, chamber temperature gradients, and turbulence. These are explored using a mixing limited vaporization model with an equation of state approach for thermopyhysical properties. This model is also applied to single and multi-component surrogates. Results include the development of the combustion research facility and validated thermodynamic state generation procedure. The developed equation of state approach provides application for improving surrogate fuels, both single and multi-component, in terms of diesel spray liquid length, with knowledge of only critical fuel properties. Experimental studies are coupled with modeling incorporating improved thermodynamic non-ideal gas and fuel

Effects of carbonation on the mineral composition of cement kiln dust

Due to their relatively high calcium oxide content, industrial mineral oxide wastes are potential candidates for mineral sequestration of carbon dioxide (CO2). Cement kiln dust (CKD), a byproduct of cement manufacturing contains 20-60% CaO making it a possible candidate for CO2 sequestration. In this study, three types of CKD are characterized, before and after carbonation, using environmental scanning electron microscopy and energy dispersive x-ray microanalysis to determine the mineralogical and morphological changes occurring due to carbonation. The reactants, products, and precipitation mechanisms were investigated to enhance understanding of the governing processes and allow better utilization of CKD for CO2 sequestration. The results of multiple independent analyses confirmed the formation of CaCO3 during carbonation. Examinations of the reaction pathways found that CaO and calcium hydroxide (Ca(OH)2) were the major reactants. Three types of CaCO3 precipitation mechanisms were observed: (1) diffusion of CO2 into Ca(OH)2 particles causing precipitation in the pores of the particle and the growth of a CaCO3 ring from the outside inward, (2) precipitation onto existing particles, and (3) precipitation from aqueous solution. The growth of a CaCO3 ring on the outside of a particle may slow further diffusion of CO2 into a particle slowing iv the overall sequestration rate. Additionally, changes caused by carbonation in the solubility of trace metals were studied by mixing pre- and post-carbonated CKD with water and analyzing the solution using inductively coupled plasma mass spectrometry. Decreases in the leaching of chromium, lead, and copper were observed, and is an incentive for use of CKD for CO2 sequestration. Equilibrium modeling using PHREEQC confirmed that CaO and Ca(OH)2 would carbonate readily and form CaCO3.

Studies of spatial clustering of inertial particles in turbulence

We present studies of the spatial clustering of inertial particles embedded in turbulent flow. A major part of the thesis is experimental, involving the technique of Phase Doppler Interferometry (PDI). The thesis also includes significant amount of simulation studies and some theoretical considerations. We describe the details of PDI and explain why it is suitable for study of particle clustering in turbulent flow with a strong mean velocity. We introduce the concept of the radial distribution function (RDF) as our chosen way of quantifying inertial particle clustering and present some original works on foundational and practical considerations related to it. These include methods of treating finite sampling size, interpretation of the magnitude of RDF and the possibility of isolating RDF signature of inertial clustering from that of large scale mixing. In experimental work, we used the PDI to observe clustering of water droplets in a turbulent wind tunnel. From that we present, in the form of a published paper, evidence of dynamical similarity (Stokes number similarity) of inertial particle clustering together with other results in qualitative agreement with available theoretical prediction and simulation results. We next show detailed quantitative comparisons of results from our experiments, direct-numerical-simulation (DNS) and theory. Very promising agreement was found for like-sized particles (mono-disperse). Theory is found to be incorrect regarding clustering of different-sized particles and we propose a empirical correction based on the DNS and experimental results. Besides this, we also discovered a few interesting characteristics of inertial clustering. Firstly, through observations, we found an intriguing possibility for modeling the RDF arising from inertial clustering that has only one (sensitive) parameter. We also found that clustering becomes saturated at high Reynolds number.

LATTICE BOLTZMANN METHOD AND CELLULAR AUTOMATA SIMULATION OF PARTICLE MOTION AND DEPOSITION IN 2-D CASE

This technical report discusses the application of the Lattice Boltzmann Method (LBM) and Cellular Automata (CA) simulation in fluid flow and particle deposition. The current work focuses on incompressible flow simulation passing cylinders, in which we incorporate the LBM D2Q9 and CA techniques to simulate the fluid flow and particle loading respectively. For the LBM part, the theories of boundary conditions are studied and verified using the Poiseuille flow test. For the CA part, several models regarding simulation of particles are explained. And a new Digital Differential Analyzer (DDA) algorithm is introduced to simulate particle motion in the Boolean model. The numerical results are compared with a previous probability velocity model by Masselot [Masselot 2000], which shows a satisfactory result.

Study of particle movement in the nearshore region of Lake Superior with radioisotope tracers

Biogeochemical processes in the coastal region, including the coastal area of the Great Lakes, are of great importance due to the complex physical, chemical and biological characteristics that differ from those on either the adjoining land or open water systems. Particle-reactive radioisotopes, both naturally occurring (210Pb, 210Po and 7Be) and man-made (137Cs), have proven to be useful tracers for these processes in many systems. However, a systematic isotope study on the northwest coast of the Keweenaw Peninsula in Lake Superior has not yet been performed. In this dissertation research, field sampling, laboratory measurements and numerical modeling were conducted to understand the biogeochemistry of the radioisotope tracers and some particulate-related coastal processes. In the first part of the dissertation, radioisotope activities of 210Po and 210Pb in a variability of samples (dissolved, suspended particle, sediment trap materials, surficial sediment) were measured. A completed picture of the distribution and disequilibrium of this pair of isotopes was drawn. The application of a simple box model utilizing these field observations reveals short isotope residence times in the water column and a significant contribution of sediment resuspension (for both particles and isotopes). The results imply a highly dynamic coastal region. In the second part of this dissertation, this conclusion is examined further. Based on intensive sediment coring, the spatial distribution of isotope inventories (mainly 210Pb, 137Cs and 7Be) in the nearshore region was determined. Isotope-based focusing factors categorized most of the sampling sites as non- or temporary depositional zones. A twodimensional steady-state box-in-series model was developed and applied to individual transects with the 210Pb inventories as model input. The modeling framework included both water column and upper sediments down to the depth of unsupported 210Pb penetration. The model was used to predict isotope residence times and cross-margin fluxes of sediments and isotopes at different locations along each transect. The time scale for sediment focusing from the nearshore to offshore regions of the transect was on the order of 10 years. The possibility of sediment longshore movement was indicated by high inventory ratios of 137Cs: 210Pb. Local deposition of fine particles, including fresh organic carbon, may explain the observed distribution of benthic organisms such as Diporeia. In the last part of this dissertation, isotope tracers, 210Pb and 210Po, were coupled into a hydrodynamic model for Lake Superior. The model was modified from an existing 2-D finite difference physical-biological model which has previously been successfully applied on Lake Superior. Using the field results from part one of this dissertation as initial conditions, the model was used to predict the isotope distribution in the water column; reasonable results were achieved. The modeling experiments demonstrated the potential for using a hydrodynamic model to study radioisotope biogeochemistry in the lake, although further refinements are necessary.

Particles induce apical plasma membrane enlargement in epithelial lung cell line depending on particle surface area dose

Background Airborne particles entering the respiratory tract may interact with the apical plasma membrane (APM) of epithelial cells and enter them. Differences in the entering mechanisms of fine (between 0.1 μm and 2.5 μm) and ultrafine ( ≤ 0.1 μm) particles may be associated with different effects on the APM. Therefore, we studied particle-induced changes in APM surface area in relation to applied and intracellular particle size, surface and number. Methods Human pulmonary epithelial cells (A549 cell line) were incubated with various concentrations of different sized fluorescent polystyrene spheres without surface charge (∅ fine – 1.062 μm, ultrafine – 0.041 μm) by submersed exposure for 24 h. APM surface area of A549 cells was estimated by design-based stereology and transmission electron microscopy. Intracellular particles were visualized and quantified by confocal laser scanning microscopy. Results Particle exposure induced an increase in APM surface area compared to negative control (p < 0.01) at the same surface area concentration of fine and ultrafine particles a finding not observed at low particle concentrations. Ultrafine particle entering was less pronounced than fine particle entering into epithelial cells, however, at the same particle surface area dose, the number of intracellular ultrafine particles was higher than that of fine particles. The number of intracellular particles showed a stronger increase for fine than for ultrafine particles at rising particle concentrations. Conclusion This study demonstrates a particle-induced enlargement of the APM surface area of a pulmonary epithelial cell line, depending on particle surface area dose. Particle uptake by epithelial cells does not seem to be responsible for this effect. We propose that direct interactions between particle surface area and cell membrane cause the enlargement of the APM.