A numerical method commonly referred to as smooth particle hydrodynamics (SPH) is implemented in two dimensions for solid mechanics in general and fracture mechanics in particular. The implementation is tested against a few analytical cases: a vibrating plate, a bending plate, a modus I crack and a modus II crack. A conclusion of these tests is that a better way of treating a shortcoming of SPH called tensile instability is needed. A study is made on the best choice of a vital parameter called the smoothing radius, and it is found that a good choice of the smoothing radius is roughly 1.5 times the initial particle spacing.

Subject of the present study was the investigation of the wake behind stepped circular cylinders. Experiments were performed using particle image velocimetry (PIV). Cylinder models with diameter ratios D/d=1, D/d=6/5 and D/d=2 were investigated at Reynolds numbers Re_{D}=2243.6, Re_{D}=12820.5 and Re_{D}=51282.1 with respect to the large diameter. Of main interest were the influences of step size and Reynolds number on the formation of vortex shedding cells. Mean flow characteristics and instantaneous captures of the flow field were analyzed in order to gain insight into vortex development in the wake. Spectral analysis was applied to obtain information on the number of forming vortex shedding cells.It was found that an increase in Re_D leads to a decrease in eddy formation length $L_f$. At the largest investigated Reynolds number, turbulence gained strong influence on the flow characteristics. An increase in stepsize leads to the development of a large and a small diameter wake with distinct eddy formation lengths L_{f,L} and L_{f,S}.Due to a mismatch in vortex shedding frequencies in the two wakes, vortices that come out of phase connect to subsequent vortices on the same side of the step. This occurs more frequently for a higher diameter ratio.Several vortex shedding cells were found in the wake of the step cylinders, their number increasing with the step size and differing for different Reynolds numbers.The wake behind step cylinders was discovered to be strongly three-dimensional. I order to complete the information about flow characteristics, measurements were performed in several parallel vertical and horizontal planes along the cylinder.PIV proved to be a good method for the investigation of flow features in the wake of step cylinders for the chosen model dimensions and Reynolds numbers.

Previous research has demonstrated that pool-riffle bedforms play a critical role in channel stability and ecosystem health in many natural gravel-bed channels. Although the bedform length is known to scale with channel width, no experimental research has yet isolated the effect of bedform length on pool-riffle hydrodynamics. To improve the understanding of the hydrodynamics of these bedforms so that they can be better incorporated in restoration practices, flume experiments were conducted testing the flow at seven different bedform lengths. Velocity profiles are measured in a 17 m flume with movable PVC bedforms using ultrasonic velocity profilers (UVPs). Smooth two-dimensional (no sinuosity) bedforms are used in order to isolate the key dynamics in convective acceleration and deceleration. The angle of transition between pool and riffle heights was 7°, so that permanent flow separation did not occur. Parameters calculated from the velocity and turbulence profiles include the Coles’ wake parameter (a measure of the deviation from the log law), shear stress estimated from the velocity profile, shear stress estimated from the Reynolds shear stress, and vertical velocity. From the individual velocity time series, the integral length scale and the integral time scales are also calculated. Overall, the length of riffles and pools exert a fundamental control on the distribution of flow and turbulence within a channel. In the pool, energy is dissipated both through turbulence and as the flow is redistributed to uniform flow conditions. In the riffle, kinetic energy increases as the flow velocity increases, and as the length increases, the flow moves towards a new uniform flow condition. The results start to explain the reasons behind the persistent scaling relation between width and bedform length. It can be concluded that uniform flow conditions exist at the end of the pool when the bedform length ratio is greater than approximately 1:5.0 when the riffle length is held constant, and that uniform flow conditions are no longer observed at the end of the pool when the bedform length ratio exceeds 1:7.0 when the pool length is held constant. Future research should concentrate on extending the results to include three-dimensional pool-riffle configurations, repeating bedform configurations, internal scaling parameters, and sediment transport. Ultimately, as the hydrodynamics of pool-riffle sequences are better understood, better bedform designs can be implemented in restoration projects.

No description available.

Understanding dynamics of colloidal dispersions is important for several applications ranging from coatings such as paints to growing colloidal crystals for photonic bandgap materials. The research outlined in this dissertation describes the use of Monte Carlo and Stokesian Dynamic simulations to model colloidal dispersions, and the development of theoretical expressions to quantify and predict dynamics of colloidal dispersions. The emphasis is on accurately modeling conservative, Brownian, and hydrodynamic forces to model dynamics of colloidal dispersions. In addition, we develop theoretical expressions for quantifying self-diffusion in colloids interacting via different particle-particle and particle-wall potentials. Specifically, we have used simulations to quantitatively explain the observation of anomalous attraction between like-charged colloids, develop a new criterion for percolation in attractive colloidal fluids, and validate the use of analytical expressions for quantifying diffusion in interfacial colloidal fluids. The results of this work contribute to understanding dynamics in interfacial and bulk colloidal fluids.

<p>The hydrodynamics of an agitated tank have been studied byphase-Doppler anemometry. The focus is on the impeller anddischarge region of a 45o pitched blade turbine (PBT). Thestudy includes agitation of pure water as well as of a dilutesuspension of process particles. A three-dimensionalphase-Doppler anemometer is used to measure local,instantaneous, three-dimensional velocities of the fluid and ofthe suspended particles. A shaft encoding technique is used toresolve the turbulent fluctuations from the periodic velocityfluctuation due to the impeller blades, and to provide moredetailed information about the variations relative to theimpeller blade. Velocity bias is corrected for by the total3-dimensional velocity.</p><p>The mean flow field, the fluctuating velocities, and thecomplete Reynolds stress tensor, are reported for the liquidphase flow. The periodic fluctuations in the flow that aregenerated by the impeller blades are eliminated in theexamination of the turbulence. The anisotropy of the turbulenceis assessed by the invariants of the anisotropy tensor. Thetrailing vortex structure is demonstrated to be associated withhigh kinetic energy and strong anisotropy of the turbulence.The vortex is still observable 130-140 degrees behind theblade. It gradually moves down from the impeller blade but thelocation in radial direction remains essentially unchanged. Theinfluence of the periodic fluctuations is examined and it isshown that the turbulence appears more isotropic when theperiodic fluctuations are not eliminated.</p><p>The solid particle concentration is low below the impellerand is high above the impeller tip. The particles diverge fromthe liquid flow mean direction, especially below the agitatorclose to the tip where the strongest turbulence is found.Periodic fluctuations in the particle concentration relate tothe variations found in the angle-resolved mean velocity andfluctuating velocity. The ratio of the maximum to the minimumconcentration is about 2.0 in the present study.</p><p>The baffles influence on the conditions in the impellerregion, and this influence can be observed on the fluid meanvelocity field, the angle-resolved velocities, the kineticenergy, and on the behavior of larger process particles. In theimpeller region the highest kinetic energies are about 15%higher upstream of the baffle than at the middle plane betweenthe baffles. The highest energy level in the middle plane isactually the lowest value and is therefore not representativewhen rotation symmetry is assumed.</p><p>Local energy dissipation rates have been investigated, andthe integration of the local energy dissipation rates overdifferent control volumes has been compared with macroscopicenergy balance calculations. The discrepancy is significant.Different reasons have been analyzed and recommendations forfurther investigation are given. I</p><p>n the outflow region there is a significant variation alsoin the direction of the instantaneous velocity, which may leadto direction bias in the case of non-spherical measurementvolume. In order to account for this direction bias, amathematical model is developed to estimate the projected areaof the measurement volume in LDA or PDA. It is shown that theprojected area variation can lead to a significant directionbias in determination of time averaged values and localparticle concentration in a highly turbulent stirred tank flow.This bias is however negligible for an orthogonal optical setup, as is used in the present study.</p><p><b>Keywords:</b>Hydrodynamics, phase-Doppler anemometer,suspension, pitched-blade turbine, anisotropy, turbulence,Reynolds stresses, trailing vortex, kinetic energy, stirredtank</p>

Thèse de doctorat : Sciences : Paris, Faculté des sciences : 1867. / Titre provenant de l'écran-titre.

Thesis (Ph.D.)--Memorial University of Newfoundland, 2001. / Bibliography: leaves 187-194.

Thesis (Ph. D.)--University of Washington, 2003. / Vita. Includes bibliographical references (leaves 73-77).

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2003. / Includes bibliographical references (leaves 108-109). Also available in electronic version. Access restricted to campus users.