This thesis describes experimental studies of colloidal particles dispersed in solvents which themselves have phase transitions. One common definition of soft matter is: a material characterized by a mesoscopic length scale. This length scale is, for example, the colloid size or the ordered domain size. Here we combine a complex host with one characteristic length scale with dispersed particles that have a different size. It might be anticipated that new behaviour will occur. Two limits of particle characteristics are probed: the case of dilute sterically-stabilized particles and the case of a weak gel of attractive particles. The two systems are polymer particles dispersed in a phaseseparating microemulsion and silica nanoparticles dispersed in a low molecular weight liquid crystal. In each system a temperature driven phase transition plays a crucial role. In the microemulsion case we observe how transitional and pre-transitional phenomena create effective interactions between particles and how new behaviour emerges in the host solvent in the late stage of the phase separation. We show that the pre-transitional clustering of the PMMA particles is due to an adsorbed layer of dodecane. Subsequently heterogeneous nucleation of the gas phase is seen. After phase separation has occurred in off-critical samples the particles remain in either the continuous or dispersed phase depending on the original microemulsion composition. In the late stage of the phaseseparation the coalescence and coarsening behaviour changes significantly, after more material exchange between the phases has taken place. This behaviour is reminiscent of viscoelastic phase separation in polymer based samples. In the liquid crystal case we discover the anisotropy of the liquid crystal persists over large length scales and modifies the local dynamics of the gel. Using electron microscopy and scattering techniques we demonstrate that the silica embedded in the liquid crystal forms agglomerates with a fractal structure. Rheological characterization demonstrates that the resulting composite is a gel. Investigation of the composite’s local dynamics using x-ray photon correlation spectroscopy shows anisotropy and intermittency in the dynamics on significant length scales. In both systems we have studied new behaviour seen due to the influence of one component on the dynamic characteristics of the other The pre- and post- phase transition phenomena are only crucial in the microemulsion case where the particles have purely repulsive interactions. Our results illustrate the subtle balances that occur in soft composite systems.

Frozen soil has a major effect in many hydrologic processes, and its impacts are difficult to predict. A prime example is flood forecasting during spring snowmelt within the Canadian Prairies. One key driver for the extent of flooding is the antecedent soil moisture and the possibility for water to infiltrate into (partly) frozen soils. Therefore, these situations are crucial for accurate flood prediction at every spring. The main objective of this study was to evaluate the water flow and heat transport within available hydrological models to predict the impact of frozen and partly frozen soil on infiltration and percolation. A standardized data set was developed for water flow and heat transport into (partial) frozen soil by laboratory experiments using fine sand within a one-dimensional (1-D) soil column. A 1-D soil column having a length of 107 cm and diameter of 35.6 cm was built and equipped with insulation to limit heat exchange only through the soil surface. A data logger collected the moisture content and temperature by five FDR sensors which have been installed at a distance of 15 cm from each other. During the experiments, temperature, soil moisture, and percolated water was observed at different freezing conditions (-5°C, -10°C, and -15°C) as well as at thawing conditions when the air temperature was increased to +5°C. Distribution of soil moisture and soil temperature in the soil column was plotted for the experimental data over the freezing and thawing period. As some of the water in the soil begins to freeze, a decrease in water content was observed with a sudden increase in soil temperature near 0°C or slightly below of 0°C. This was, in fact, only a decrease in unfrozen water, not a decrease in total water content and was caused by the latent heat during freezing. Soil temperature showed noticeable differences at the top and the bottom of soil column during the change of state of water. The heat flux at the lower soil column was strongly limited due to iii the overlying soil. Thus, the soil temperature at the lowest sensors stayed in a freezing condition over several days and was not changing the temperature due to the latent heat which was released during the freezing process. Significant variation in soil moisture content was found between the top and the bottom of the soil column at the starting of the thawing period. However, with increasing temperature, the lower depth of the soil column showed higher moisture content as the soil was enriched with moisture with higher transmission rate due to the release of heat by soil particles during the thawing cycle. The soil system did not remain in the isothermal state during the thawing cycle. Although gravitational gradient was mainly responsible for the infiltration rate into the partially frozen soil, the distribution of moisture was greatly influenced by the temperature gradient. Vadose zone modeling using HYDRUS-1D was applied to the data set. Numerical results of the modeling were calibrated using the experimental results. It showed that the newly developed benchmark data set were useful for the validation of numerical models. The use of such a validated freezing and thawing module implemented into larger scale hydrologic models will directly reduce the prediction uncertainty during flood forecasting. Moreover, these benchmark data sets will be useful for the validation of numerical models and for developing scientific knowledge to suggest potential code variations or new code development in numerical models. / February 2017

Construction practices are constantly evolving in order to adapt to physical locations and economic conditions. These adaptations may result in more cost-effective designs, but may also come at a cost of strength. In masonry shear walls, it is becoming more common to reduce the amount of grouting from every cell to only those with reinforcement, a practice known as partial-grouting. Partially-grouted masonry responds differently and in a more complex matter to lateral loads as compared to fully-grouted masonry. The response is made even more complex by wall discontinuities in the form of openings. The main objective of this study is to validate the strut-and-tie procedure for the in-plane lateral strength prediction of partially-grouted, multistory, reinforced concrete masonry walls with openings. The research included testing six three story, half-scale masonry shear walls. Half of the walls had door openings while the other half had window openings. The configurations were selected to represent typical walls in multi-story buildings. The measured lateral strength was compared to estimations from the equations in the US masonry code and to those from an equivalent truss model and a strut-and-tie model. The results show that the U.S. masonry code equations over predicts while the equivalent truss model under predicts the lateral strength of the walls. The results further show that the strut-and-tie model is the most accurate method for lateral strength prediction and is able to account for wall openings and partial-grouting.

Partially hydrogenated oil (PHO) has no longer GRAS status. However, PHO is one of the important ingredients in bakery and confectionary industry and therefore the food industry is seeking for an alternative fat to replace PHO. Waxes have shown promise to fulfill that demand because of its easy availability and cheap in price. Waxes with high melting points (> 40 °C) help in the crystallization process when mixed with low melting point oils. A crystalline network is formed in this wax/oil crystallization process where liquid oil is entrapped in wax crystal network. A new material is formed which is neither completely solid nor completely liquid; it’s called semisolid material. This wax/oil semisolid material is formed physically; there are no chemical processes or treatments involved. This material has a potential use in the lipid industry due to its resemblance to the properties of commercial margarine or similar lipids. BW has shown softer crystalline network formation compared to SFW and RBW. It is understood that presence of higher wax ester in SFW and RBW leads to stronger crystalline material formation. Blending waxes of different chemical composition (e.g. BW: wax ester, hydrocarbon, fatty acids, di-esters, hydroxyl esters. RBW: 100% wax ester) shows differences in physical characteristics at different blending proportions. HIU technology helps in delaying phase separation of crystals in low concentration (0.5 and 1% wt. basis) of wax/oil system. Our overall wax crystallization study has shown that there are different physical characteristics of wax/oil semi-solid system based on different parameters and processing conditions such as wax concentration, wax and oil type, cooling rate, storage temperature, high intensity ultrasound. The hypothesis of this dissertation is that chemical composition of waxes and vegetable oils and also processing conditions affect wax crystallization and physical properties of wax/oil materials.

The LYM property of a finite standard graded poset is one of the central notions in Sperner theory. It is known that the product of two finite standard graded posets satisfying the LYM properties may not have the LYM property again. In 1974, Harper proved that if two finite standard graded posets satisfying the LYM properties also satisfy rank logarithmic concavities, then their product also satisfies these two properties. However, Harper's proof is rather non-intuitive. Giving a natural proof of Harper's theorem is one of the goals of this thesis. The main new result of this thesis is a characterization of rank-finite standard graded LYM posets that satisfy rank logarithmic concavities. With this characterization theorem, we are able to give a new, natural proof of Harper's theorem. In fact, we prove a strengthened version of Harper's theorem by weakening the finiteness condition to the rank-finiteness condition. We present some interesting applications of the main characterization theorem. We also give a brief history of Sperner theory, and summarize all the ingredients we need for the main theorem and its applications, including a new equivalent condition for the LYM property that is a key for proving our main theorem.

The LYM property of a finite standard graded poset is one of the central notions in Sperner theory. It is known that the product of two finite standard graded posets satisfying the LYM properties may not have the LYM property again. In 1974, Harper proved that if two finite standard graded posets satisfying the LYM properties also satisfy rank logarithmic concavities, then their product also satisfies these two properties. However, Harper's proof is rather non-intuitive. Giving a natural proof of Harper's theorem is one of the goals of this thesis. The main new result of this thesis is a characterization of rank-finite standard graded LYM posets that satisfy rank logarithmic concavities. With this characterization theorem, we are able to give a new, natural proof of Harper's theorem. In fact, we prove a strengthened version of Harper's theorem by weakening the finiteness condition to the rank-finiteness condition. We present some interesting applications of the main characterization theorem. We also give a brief history of Sperner theory, and summarize all the ingredients we need for the main theorem and its applications, including a new equivalent condition for the LYM property that is a key for proving our main theorem.

Shared objects are a key component in today's large distributed systems. Linearizability is a popular consistency condition for such shared objects which gives the illusion of sequential execution of operations. The time bound of an operation is the worst-case time complexity from the operation invocation to its response. Some time bounds have been proved for certain operations on linearizable shared objects in partially synchronous systems but there are some gaps between time upper bound and lower bound for each operation. In this work, the goal is to narrow or eliminate the gaps and find optimally fast implementations. To reach this goal, we prove larger lower bounds and show smaller upper bounds (compared to 2d for all operations in previous folklore implementations) by proposing an implementation for a shared object with an arbitrary data type in distributed systems of n processes in which every message delay is bounded within [d-u, d] and the maximum skew between processes' clocks is epsilon. Considering any operation for which there exist two instances such that individually, each instance is legal but in sequence they are not, we prove a lower bound of d + min{epsilon, u, d/3}, improving from d, and show this bound is tight when epsilon < d/3 and epsilon < u. Considering any operation for which there exist k instances such that each instance separately is legal and any sequence of them is legal, but the state of the object is different after different sequences, we prove a lower bound of (1-1/k)u, improving from u/2, and show this bound is tight when k = n. A pure mutator only modifies the object but does not return anything about the object. A pure accessor does not modify the object. For a pure mutator OP1 and a pure accessor OP2, if given a set of instances of OP1, the state of the object reflects the order in which the instances occur and an instance of OP2 can detect whether an instance of OP1 occurs, we prove the sum of the time bound for OP1 and OP2 is at least d + min{epsilon, u, d/3}, improving from d. The upper bound is d + 2*epsilon from our implementation.

Semiparametric models play important roles in the &#64257;eld of biological statistics. In this dissertation, two types of semiparametic models are to be studied. One is the partially linear model, where the parametric part is a linear function. We are to investigate the two common estimation methods for the partially linear models when the data is correlated — longitudinal or clustered. The other is a semiparametric model where a latent covariate is incorporated in a mixed effects model. We will propose a semiparametric approach for estimation of this model and apply it to the study on colon carcinogenesis. First, we study the pro&#64257;lekernel and back&#64257;tting methods in partially linear models for clustered/longitudinal data. For independent data, despite the potential rootn inconsistency of the back&#64257;tting estimator noted by Rice (1986), the two estimators have the same asymptotic variance matrix as shown by Opsomer and Ruppert (1999). In this work, theoretical comparisons of the two estimators for multivariate responses are investigated. We show that, for correlated data, back&#64257;tting often produces a larger asymptotic variance than the pro&#64257;lekernel method; that is, in addition to its bias problem, the back&#64257;tting estimator does not have the same asymptotic ef&#64257;ciency as the pro&#64257;lekernel estimator when data is correlated. Consequently, the common practice of using the back&#64257;tting method to compute pro&#64257;lekernel estimates is no longer advised. We illustrate this in detail by following Zeger and Diggle (1994), Lin and Carroll (2001) with a working independence covariance structure for nonparametric estimation and a correlated covariance structure for parametric estimation. Numerical performance of the two estimators is investigated through a simulation study. Their application to an ophthalmology dataset is also described. Next, we study a mixed effects model where the main response and covariate variables are linked through the positions where they are measured. But for technical reasons, they are not measured at the same positions. We propose a semiparametric approach for this misaligned measurements problem and derive the asymptotic properties of the semiparametric estimators under reasonable conditions. An application of the semiparametric method to a colon carcinogenesis study is provided. We &#64257;nd that, as compared with the corn oil supplemented diet, &#64257;sh oil supplemented diet tends to inhibit the increment of bcl2 (oncogene) gene expression in rats when the amount of DNA damage increases, and thus promotes apoptosis.

The objective of the study is to investigate the capability of PANS (Partially Averaged Navier-Stokes Simulation) model over a wide range of Reynolds numbers and flow physics. In this regard, numerical simulations of turbulent flow past a circular cylinder are performed at ReD 140,000 and ReD 3900 using the PANS model. The high Reynolds number PANS results are compared with experimental results from Cantwell and Coles, Large Eddy Simulation results from Breuer, and Detached Eddy Simulation results from Travin et al. Low Reynolds number PANS results are compared with experimental results from Ong and Wallace and Large Eddy Simulation results from Breuer. The effects of the various PANS parameters (fk, f&#949;, &#963;ku, &#963;&#949;u) on the ability to capture turbulence physics at various Reynolds numbers are studied. It is confirmed, as previously predicted from theoretical considerations that: (i) for high Reynolds number flow f&#949; = 1 and &#963;ku = &#963;k × fk2 / f&#949; are most appropriate; and (ii) for low Reynolds number flow f&#949; = fk and &#963;ku = &#963;k are most suitable. These choices for the parameters stem from the fact that there is no clear separation of scales between the energy scales and the dissipation scales at low Reynolds number unlike in the high Reynolds number where there is a clear separation of scales between the energy containing scales and the dissipation scales. Also, in both cases it is found that decreasing fk leads to improved accuracy in predicting the flow statistics.

The multilevel model (MLM) is easily parameterized to handle partially clustered data (see, for example, Baldwin, Bauer, Stice, & Rohde, 2011). The current study evaluated the performance of this model under various departures from underlying assumptions, including assumptions of normally distributed and homoscedastic Level 1 residuals. Two estimating models – one assuming homoscedasticity, the other allowing for the estimation of unique Level 1 variance components – were compared. Results from a Monte Carlo simulation suggest that the MLM for partially clustered data yields consistently unbiased parameter estimates, except for an underestimation of the Level 2 variance component under heteroscedastic generating conditions. However, this negative parameter bias desisted when the MLM allowed for Level 1 heteroscedasticity. Standard errors for variance component estimates at both levels were underestimated in the presence of nonnormally distributed Level 1 residuals. A discussion of results, as well as suggestions for future research, is provided. / text