Content area
Abstract
This thesis consists of two parts. In part I, turbulence statistics and flow structure are studied in a fully-developed rough-wall channel flow Reτ in the 3520–5360 range. A facility with optically index-matched fluid enables applying Particle Image Velocimetry (PIV) very near a rough surface composed of closely-packed pyramidal elements with height k=0.46 mm and wavelength λ=3.2 mm. The flow and roughness settings achieve "well-characterized" (Jiménez 2004) flow conditions with h/k≈50 (h is half channel height) and k+=60∼100, which is suitable for investigating wall-similarity hypothesis. By performing planar and time-resolved PIV measurements of streamwise-wall-normal (x-y)and streamwise-spanwise ( x-z) planes at multiple resolutions, we examine the turbulence statistics and dynamics of flow structure from within the roughness to the channel centerline. The results show that the spatial variations in mean flow, Reynolds stresses, as well as turbulent kinetic energy (TKE) production and dissipation rates are confined to 2k above the surface, the region referred to as the roughness sublayer in the present context. Moreover, all the spatially-averaged Reynolds stress components have local maxima at slightly higher elevations, but the streamwise-normal component increases rapidly within 1k away from the wall, peaking at the top of the pyramids. Similarly, the TKE production and dissipation rates also peak near the wall, and their ratio has a maximum below 1k above the roughness but decreases to below one in the outer layer. Wall-normal turbulence transport is significant only close to the roughness. The spatial turbulent energy and shear spectra show an increasing contribution of large-scale motions, and diminishing role of small scale ones with increasing distance from the wall. However, as the spectra steepen at low wavenumbers, they flatten and develop bumps in wavenumbers corresponding to 1−3k. In compensated spectra, these bumps appear as additional maxima at scales corresponding to 15–30 times the Kolmogorov scales, i.e. within the dissipation range. Following this observation, we decompose the turbulence to large (>λ), intermediate (3–6k), roughness (1–3k) and small (<k) scales, where k and λ are roughness height and wavelength, respectively. With decreasing distance from the wall, there is a remarkable increase in the 'non-local' SGS energy flux directly from large to small scale, and the fraction of turbulence dissipated by roughness scale eddies.
In part II, the grazing behaviors of calanoid copepod Acartia tonsa are investigated. Our objective is to quantify the response of marine copepods to varying diets, in particular when they are exposed to toxic dinoflagellates involved with harmful algal blooms (HABs). The conjectured role of toxins as a grazing deterrent is investigated by implementing digital holographic cinematography to compare the behavior of free-swimming Acartia tonsa on the nutritional diet of Storeatula major to that occurring during exposure to toxic and non-toxic strains of Karenia brevis and Karlodinium veneficum. Statistical analysis of the duration of the copepod's feeding appendage beating enables us to distinguish between two beating modes with lognormal distributions: "sampling beating" that has short durations (<100 ms) and involves little fluid entrainment, and "grazing beating" that persists for longer periods (up to 1200 ms) and generates feeding currents. Without prey, A. tonsa only samples the water at low frequency, and upon introduction of desired food, it increases its sampling time moderately and the grazing substantially. On mono algal diets of either toxic dinoflagellate, the sampling frequencies are high, but the grazing remains low, and the copepod demonstrates aversion to both. In mixtures of S. major and toxic K. brevis, the sampling and grazing diminish rapidly, presumably due to neurological effects of consuming brevetoxins during while trying to feed on S. major. In contrast, on mixtures of toxic K. veneficum, both modes persist, indicating that intake of karlotoxins does not inhibit the copepod's grazing ability. Possible reasons for the lognormal distributions of duration, and the ecological impact of the different responses to toxins are discussed. (Abstract shortened by UMI.)
