Browsing by Subject "Coalescence"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Euler-Lagrangian simulations of turbulent bubbly flow.
    (2011-03) Mattson, Michael David
    A novel one-way coupled Euler-Lagrangian approach, including bubble-bubble collisions, coalescence and variable bubble radius, was developed in the context of simulating large numbers of cavitating bubbles in complex geometries using direct numerical simulation (DNS) and large-eddy simulation (LES). This dissertation i) describes the development of the Euler-Lagrangian approach, ii) outlines the novel bubble coalescence model derived for this approach and iii) describes simulations performed of bubble migration in a turbulent boundary layer, bubble coalescence in a turbulent pipe ow and cavitation inception in turbulent flow over a cavity. The coalescence model uses a hard-sphere collision model is used and determines coalescence stochastically. The probability of coalescence is computed from a ratio of coalescence timescales, which are dynamically determined from the simulation. Coalescence in a bubbly, turbulent pipe ow (Re#28; = 1920) in microgravity was simulated with conditions similar to experiments by Colin et al. [1] and excellent agreement of bubble size distribution was obtained. With increasing downstream distance, the number density of bubbles decreases due to coalescence and the average probability of coalescence decreases due to an increase in overall bubble size. The Euler-Lagrangian approach was used to simulate bubble migration in a turbulent boundary layer (420 < Re#18; < 1800). Simulation parameters were chosen to match Sanders et al. [2], although the Reynolds number of the simulation is lower than the experiment. The simulations show that bubbles disperse away from the wall as observed experimentally. Mean bubble diffusion and profiles of bubble concentration are found to be similar to the passive scalar results, except very near the wall. The carrier-fluid acceleration was found to be the reason for moving the bubbles away from the wall. The one-way coupled Euler-Lagrangian approach was applied to simulate the experiment of cavitating turbulent ow over a cavity by Liu and Katz [3]. The classical Rayleigh-Plesset equation is integrated using adaptive time-stepping to accurately and efficiently solve for the change of the bubble radius over time. The one-way coupled Euler-Lagrangian model predicts cavitation inception at the trailing edge of the cavity and also in the vortices shed from the leading edge, in qualitative agreement with experiment.
  • Thumbnail Image
    Item
    Phylogeography of palearctic birds using mulyilocus coalescence analyses
    (2012-07) Hung, Chih-Ming
    The phylogeography of three widespread Palearctic passerines were characterized in this dissertation based on sequence data from mitochondrial and nuclear genes, which were analyzed under a framework of the coalescence. The following questions were addressed: (1) does gene history reflect species history, (2) does natural selection on mitochondrial DNA (mtDNA) obscure phylogeographic inference, and (3) do the early stages of speciation reveal ecological niche divergence. In chapter one, phylogeographic histories based on mtDNA and 13 nuclear genes of Eurasian nuthatches (Sitta europaea) were compared to address an ongoing debate over the value of mtDNA in phylogeography. Both mtDNA and multilocus nuclear data recovered the same three clades. The results suggested that mtDNA is efficient in discovering phylogeographic pattern due to its fast coalescence rate; whereas, multiple (nuclear) genes are required to quantify process parameters such as effective population size, gene flow and divergence time. In chapter two, I devised novel methods based on coalescent simulations to discover whether natural selection has influenced the mitochondrial genome of two Old World flycatcher species (Ficedula parva and F. albicilla). The simulations were based on the estimated demographic history using 18 nuclear genes, which suggested that the two sister species diverged around three million years ago and that F. albicilla, but not F. parva, experienced a recent population expansion. My analyses showed that population bottlenecks alone could not fully explain the strikingly low variation in the mtDNA data, and I concluded that the mtDNA patterns were affected by natural selection. Thus, interpreting the phylogeographic history based solely on mtDNA can be misleading. Chapter three involved coalescence and simulation approaches based on mtDNA and a Z-linked gene to explore the early stages of evolutionary divergence in the common rosefinch (Carpodacus erythrinus). The third chapter also involved an assessment of ecological niche divergence and the results showed no evidence of ecological divergence at the early stages of diversification among three rosefinch lineages. This dissertation demonstrates that the incorporation of multiple genes, coalescence theory, and analytic approaches from other fields (e.g., ecological niche modeling) can provide fresh insights into phylogeographic history of species.