The work being presented here concentrates on characterizing flow scales of turbulent jets in the near field. Experiments were carried out with water jets and immiscible silicone oil jets of two viscosities submerged in a water tank. The jet Reynolds numbers are in the range of Re ~ 4500 - 50000 for homogeneous water jets and Re ~ 3500 - 27000 for silicone oil jets in water. To observe the turbulent/non-turbulent interface, the jet fluids are made visible by doping with fluorescent dye and excitation with directional illumination. The jet interfaces are continuous and convoluted for water jets, whereas the interfaces of the oil jets are convoluted and discontinuous with droplets and ligaments. Direct flow visualization, schlieren photography, shadowgraph photography and particle image velocimetry are employed as appropriate. Interfacial length scales are characterized using various image processing techniques for both water and oil jet runs. For the homogeneous water jet runs, streamwise internal length scales of the schlieren recording are investigated after applying temporal filters to isolate scales within the flow. When isolated based on their temporal signatures, observations of the homogeneous water jets show that the length scales of internal features increase with Reynolds number. As the Reynolds number increases, the temporal signatures of the features within the jet can be seen populating higher frequency modes. Droplet sizes for the immiscible jet runs are quantified using Hough transformation. Interface length scales decrease with Reynolds number and increase gradually with distance from the exit plane for a given Reynolds number. These scales are isotropic for the homogeneous water jets and exhibit a streamwise to cross-stream ratio of about 1.3 for the oil jets. Results indicate the average droplet size in the immiscible jets is determined by the interfacial surface tension, in relation to the Weber number.

Experiments are conducted to model the flow within a cerebral saccular aneurysm located at the basilar artery bifurcation. The flow phantom modeling the aneurysm consists of a nearly spherical dome located at the flow divider of a 90° bifurcation. To ensure that the experiments are physiologically relevant, a pulsatile pumping system replicates a measured basilar artery waveform at the inlet. In addition, the Reynolds and Sexl-Womersley numbers are matched with physiological conditions. The flow rate through the outlets of the bifurcation are controlled by a pair of needle valves. This control allows study of the effect of branching ratio. Three techniques are applied to characterize the flow. Flow visualization provides pathline images that are used to qualitatively define flow structures and patterns. Particle image velocimetry is used to measure the flow velocity across multiple cross sections of the flow. Post-processing of the data allows in-plane wall shear stresses to be calculated. Lastly, stereoscopic particle image velocimetry allows the 3D velocity field and in-plane wall shear stresses to be determined on an orthogonal cross section of the flow. Combining the information obtained through the experiments, the flow behavior over a single physiological waveform is characterized.

The flow characterization centers around two major features: an inlet wall jet that originates from the neck of the aneurysm and wraps around the dome and a circulation region that dominates the center volume of the dome. The motion of these structures during the waveform are captured and detailed. Wall shear stresses are found for points along the dome over the course of a waveform. In-plane maximum wall shear stress magnitudes over the entire dome vary from 6.8-18.6 dynes/cm² in the PIV planes and 13.9-31.0 dynes/cm² in the SPIV planes. Total wall shear stress magnitudes can be calculated for three points intersection points. Mean wall shear stress magnitudes at these points range from 3.28-7.94 dynes/cm² . In particular, the flow conditions that had nearly symmetric outlet flow rates had the lower WSS magnitudes of the test cases.

The primary effect of the branching ratio on the flow behavior is to alter the location and displacement of the circulation region and inlet jet. For flows with a large outlet differentials, the inlet flow shows strong direction preference entering as a wall jet. For flows with small outlet differentials, the inlet flow forms a expanding inlet jet that does not show strong direction preferences. The WSS measurements matches well with computational fluid dynamics simulations of cerebral saccular aneurysms, reinforcing the theory that lower than normal WSS values plays a role in aneurysm growth. Future experimental research will need to focus on implementing techniques that provide the complete 3D flow field and that uses a non-Newtonian working fluid to better simulate the behavior of blood.

This study describes the generation of electrolytic hydrogen bubbles in a recently proposed artificial-photosynthetic device. Bubble growth and the resulting size distributions are important inputs for predictive tools that model a device's performance factors, such as light scatter, electrolyte resistance, and volumetric gas collection. Motivated by efforts to model the performance of a micropillar-based artificial-photosynthetic device, hydrogen bubbles are produced electrochemically on a single electrode pillar. High-speed visualization and bubble-tracking algorithms are used to track the growth and record the bubble-size distributions.

The study of bubble growth defines four different modes of growth after the onset of nucleation: 1) isolated growth, 2) proximity growth, 3) surrounded growth, and 4) trailing growth. As predicted by theory, bubble diameters grow proportionally with the square root of time. Experimental results suggest that this proportionality is dependent on the mode of growth. Bubbles undergoing surrounded growth show the fastest growth rates. And when present, bubbles in trailing growth show the slowest growth rate. The former is due primarily to local microconvection currents replenishing dissolved gas at the bubble interface, and the latter is due to bubbles competing for the same gas supply. Additionally, the growth rates are observed to increase proportionally with the square root of current density as well. A formal dimensional analysis is carried out to derive the rational dependence of bubble diameters on both time, current density, and growth mode.

The study of bubble-size distribution examines the final sizes of bubbles after detachment. The size distributions of these bubbles resemble log-normal distributions, which are consistent with observations cited in literature. Experimental results reveal that most of the distributions are bimodal, where a secondary distribution of bubbles results primarily from bubble coalescence prior to detachment. Correlations are developed between the salient characteristics of the bimodal distributions and the applied current density. These correlations predict the size distributions for bubbles generated within the target current-density range of 5 to 15 mA/cm^2.

Lastly, a collection efficiency is defined to quantify how efficient is gas collection in practice. At the highest tested current density of 15 mA/cm^2, the collection efficiency is 68%. The complementary losses are due primarily to unaccounted bubbles remaining on the electrode surfaces. A loose curve fit is used to describe the progressively decreasing slope between points. This correlation provides a target current density to achieve depending on the desired collection efficiency.

These results contribute to both the fundamental-discovery and prototype-development goals behind developing a working artificial-photosynthetic device. Current density is quantified as a fundamental factor in electrolytic bubble growth. And correlations for bubble size distributions and collection efficiencies serve as inputs for modeling device performance.

An extensive sets of experiments are carried out regarding the stability characteristics of helical vortex filaments o a coaxial rotor model. A coaxial rotor model consisting of two 26-cm (10.25-in) diameter counter-rotating rotors with a root cut-out of 3.2 cm (25%) is employed in these experiments in one- and two-bladed configurations. Axial rotor separation varied from 25% of the rotor radius to one rotor radius. Rotor speed ranged from 2 to 8 rps at each rotor spacing. Dye flow visualization, particle flow visualization and particle image velocimetry are employed in order to take measurements of the flow field. For one-bladed tests, helical tip vortex filament trailed from the lower rotor blade is observed to develop strong deformations soon after its formation at the rotor spacing of a quarter rotor radius as opposed to that of upper rotor. This deformation is less pronounced at the rotor spacing of 0:8R. Upper and lower rotor vortices are seen to develop long-wave vortex pairing instabilities where they orbit around each other as they travel downstream, particularly at smaller rotor spacing, which is also confirmed in PIV results. Successive turns of the upper rotor helices appear to be stable as opposed to unstable nature of lower rotor helix in both rotor separations. At H=R = 0:8, the upper rotor wake contraction is markedly higher than that of the lower rotor. Vortex filaments develop short-wave instabilities at 2 & 4 rps just immediately after their inceptions in both rotor separations. In two-bladed configurations, in all cases tip vortices trailed from the lower rotor experience extreme distortions, resulting in the loss of their orderly helical shapes, which is a clear manifestation of the effect of the upper rotor on the lower one. At smaller rotor spacings up to 50% radius, lower rotor filaments develop long-wave vortex pairing instabilities in similar ways to what is found in single rotor wakes although they differ to the extent lower vortex filaments undergo deformations. At this rotor spacing range, upper rotor filaments does not have sufficient space between the rotors to develop mutual interactions before they reach the plane of the lower rotor and interact with lower rotor blades, i.e. interrotor blade-vortex interactions (BVIs). As the rotor spacing is further increased up to a rotor radius, upper rotor filaments are seen to develop long-wave pairing instabilities similar to what is observed in the wakes of single rotor runs. As in the one-bladed configurations, upper rotor filaments possess a more stable nature whereas those from lower rotor becomes unstable within first rotor revolution beneath the lower rotor. At rotor spacing of 50% radius, hairpin vortices are observed to form beneath the lower rotor during the long-wave instability mode developed by the lower rotor filaments where streamwise elongation of the vortex filament from a particular blade of the lower rotor in each rotor revolution is observed. At other rotor spacings, similar deformation of the vortex filaments in the streamwise direction is seen to occur however distinct hairpin vortex structures are only seen at the rotor spacing of 0:5R. Upper rotor wakes contract to more inward than those of the lower rotor wakes for all the rotor separation distances tested, which is also confirmed from the PIV measurements and particle visualization results. Increasing rotor spacing changes the wake contraction of the upper rotor at the plane of the lower rotor where it reaches a more fully-developed state towards the rotor spacing of one radius. This is further confirmed from the streamline patterns from the PIV results. Regardless of the rotor spacing, short-wave instabilities formed along the helical filaments at 2 & 4 rps cases from their generation whereas at 6 & 8 rps filaments are only observed to develop these types of instabilities long after their formation and not in an explicit manner since they become superimposed with long-wave instability mode where filaments develop strong distortions beneath the lower rotor.

Part I: Vortical flows over spinning cones at angles of incidence

The aerodynamic performance of conically-shaped bodies are directly influenced by the flowsaround them. Flows over stationary cones develop symmetric trailing vortices near their leeward surface, where any induced lateral forces are neutralized; however, a cone spinning about its axis breaks the symmetry, resulting in unequal lateral forcing. To the author’s knowledge, there are no comprehensive experimental studies that characterize the asymmetric vortex systems behind spinning cones, a flow problem that has many implications on the trajectory and stability of spin-stabilized bodies. Moreover, the laminar-turbulent transition within the boundary layer has significant implications on skin friction and drag for incompressible flows; over spinning cones, transitions form as circular and spiral waves resulting from cross-flow and centrifugal instabilities. To date, the effects of large angles of incidence on the laminar-turbulent transition are not well-understood.

Vortical flows over spinning cones with half-angles 10 (deg) ≤ θc ≤ 45 (deg) at angles of incidence, 0 ≤ α ≤ 36 (deg), are visualized using a smoke-wire technique, and measurements of the flows are made using a planar particle image velocimetry technique. The Reynolds numbers for all cones are O(10,000) and the range of rotational speed ratios is 1 ≤ |S| ≤ 3. Symmetric vortex triads composed of primary, secondary, and tertiary vortices are aligned parallel with the leeward surface of nonspinning cones at sufficient angles of incidence, α ≳ θc. The triads grow in size and strength as they are continuously fed by the adjacent shear layer. Asymmetries in vortex systems over spinning cones are characterized by anti-cyclonic vortices in the counter-rotating meridian and cyclonic vortices in the co-rotating meridian. Generally, the anti-cyclonic vortices grow as they embrace the surface of the cone and are pushed in the direction of rotation. At sufficient distances from the vertex of the cone, the anticyclonic vortices burst and initiate turbulent events downstream. Contrarily, cyclonic vortices readily detach from the leeward surface and cease to grow as they are cut off from the vorticity supply; an increase in SD or α expedites the detachment of cyclonic vortices.

Distinct behaviours of the cyclonic and anti-cyclonic vortices arise for flows over the most slender cone, θc = 10 (deg). The distinction is attributed to vortex triads forming in closer vicinity to one another as the half-angle of the cone decreases. For θc = 10 (deg), cyclonic vortices are significantly influenced by the induced flows of the anti-cylonic vortices; they are pulled past the plane of symmetry into the counter-rotating meridian and smothered between the anti-cyclonic vortex triad and the surface of the cone, initiating detachment of anti-cyclonic vortices. Distinct flows are also observed over the largest half-angle cone corresponding to θc = 45 (deg). At nonzero angles of incidence, the stagnation point departs from the vertex of the cone and monotonically shifts along the windward surface. Low-speeds are measured near the vertex of the cone at zero incidence, resembling stagnation flows near the central axis of a disc. Regular vortex shedding events in the wake region are detected in the flow visualizations, a common characteristic of flows over discs and, more generally, bluff bodies.

Boundary layer instabilities are detected and studied in the smoke streakline images. Regularly spaced wave patterns and small-scale wavy rolls consistently mark streaklines near the surfaces of the cones. The rolls form near the leeward surface of spinning cones and in streaklines ingested deep within the rotating boundary layer; hence, they are likely signatures of centrifugal spiral wave instability. The rolls leave small-scale waves on detached portions of trailing vortices that deteriorate their coherency. Contrarily, angled wave patterns form on streaklines over the entire surface of the cones in both nonspinning and spinning cases; hence, they indicate signatures of cross-flow instabilities. The flow visualizations of boundary layer instabilities primarily serve as motivation for future, more comprehensive studies.

Part II: Vortical flows over a spinning disc at incidence

Flows normal to the surface of stationary discs exhibit periodic motions in their wakes attributed to vortex shedding events. The present study reveals and characterizes, for the first time, the origins of the shedding vortices near the upstream surface of a disc. Flows are studied over a disc at incidence angles between 0 and 36 (deg), both fixed and spinning with a rotational speed ratio of 2, and the Reynolds number of 27,000. A smoke-wire technique is used for flow visualization and planar particle image velocimetry for measurements near the plane of symmetry. Coherent vortical structures are observed near the upstream surface of the disc over the full range of the incidence angle. As the structures grow in size, they align themselves parallel to the surface of the disc and shed into the wake region. Two vortex shedding modes are observed. The first is dominant at low angles of incidence, up to ∼ 18 (deg). At the limiting case of normal flow, the vortical structures shed into the wake at the Strouhal number of about 0.2. At incidence beyond 18 (deg), a secondary mode becomes dominant that appears as a soliton on the vortical structures. It originates near the leading edge of the disc, traverses towards the trailing edge, and sheds into the wake region nearly twice as frequent as the first mode. The flows over spinning discs generally mimic the stationary disc flows; however, centrifugal forces affect the formation and decay of the vortical structures. Centrifugal effects leave cross-stream instability features on the vortical structures that are likely attributed to spiral wave instabilities within the boundary layer.

The vehicles in a platoon will experience transient aerodynamic forces as vehicles leave and join the platoon at various locations. A platoon of scale vehicle models is placed in a wind tunnel and measurements are made of the transient forces and moments as one of the vehicles is moved into and out of the platoon. The results from the wind tunnel experiments will allow the computer vehicle control algorithms to better predict the transient aerodynamics the vehicles in the platoon will encounter during leaving and joining maneuvers. \par Since a lane change (either leaving or joining a platoon) maneuver occurs on the same time scale as changes in the flow field, characterization of the transient forces and moments becomes of paramount importance. In these experiments, one vehicle is moved out of and into the platoon at six different accelerations to simulate lane changes at six different time scales, the longest (smallest acceleration) representing the static case and the shortest (largest acceleration) representing half of the nominal time of an actual lane change. Measurements are made with the displaced vehicle as each of the four members of the platoon. The increase in the drag on the remaining cars in the platoon as one vehicle leaves confirms that the close spacing between vehicles in a platoon results in streamlined flow. All of the cars experience side forces and yawing moments resulting from the interrupted flow field caused by the displaced car. These effects are more pronounced on a platoon of boxes, bluffer shapes. These forces and moments are summarized here as nondimensional coefficients plotted with the nondimensional lateral displacement of the maneuvering vehicle.

This dissertation presents a design for a novel water intake mechanism for a buoy-type ocean wave energy converter (WEC). These renewable energy-harvesting devices float in the open sea and are set into vertical oscillatory motion by incident ocean waves, extracting energy from the relative motion of two or more component bodies. Energy extraction in a WEC can be performed by a variety of power takeoff systems (PTO), such as hydraulic, overhead turbines, or linear magnet generators. The work presented in this dissertation is concerned with improving the power harnessing capabilities of buoy-type WECs. To this end, we propose a novel mass modulation scheme and present designs for the associated water intake mechanism. The water intake mechanism traps and releases surrounding water as needed, thereby leading to a variation of the system's mass. The mass modulation is designed to improve the power harnessing potential of a WEC by varying the system mass at a rate of twice the frequency of the incident ocean waves.

To investigate the feasibility of the mass modulation scheme, a simple numerical model for a WEC equipped with the water intake mechanism is proposed and analyzed. Of particular interest is the relationship between mass modulation and energy harvesting, as well as the stability implications for the WEC of such a mass variation. The motions of the system have been studied in response to a spectrum of harmonic excitation and the results were catalogued. Numerical simulations have also been used to demonstrate that when applied correctly, mass modulation can lead to a significant increase in system response and power harnessing potential of a WEC.

A scale prototype of a WEC with the water intake mechanism has been constructed and tested in a wave tank to prove the concept of the proposed mass modulation scheme. The results presented in this dissertation show that a WEC fitted with the water intake mechanism exhibits a higher vertical velocity response, compared to a non-mass modulated WEC. In addition, the relative velocity between the WEC and a secondary float can also be increased using the mass modulation. These velocity increases can lead to improvements in the power harnessing potential of the WEC.

This experimental study of turbulent boundary layer separation from a contoured backward facingramp demonstrates that a spray-on superhydrophobic coating modifies the location of turbulent boundary layer separation. A custom water tunnel was designed and constructed to study turbulent boundary layer separation from a backward facing ramp located downstream of a flat plate. Particle image velocimetry was used to measure the velocity field and identify the onset of turbulent boundary layer separation. The flow rate within the water tunnel was adjusted to study the separation of turbulent boundary layers at Reynolds numbers based on the plate length of 500,000 to 6,000,000.

Experiments on a smooth baseline surface showed that the turbulent boundary layer remained fullyattached along the ramp at low flow rates. The boundary layer detached as the fluid velocity was increased, and the location of the onset of separation advanced upstream with increasing velocity. The application of a superhydrophobic coating caused the onset of separation to occur at a lower flow rate, as well as the location of separation to advance further upstream at a fixed flow rate. Analysis of the velocities in the turbulent boundary layer upstream of the ramp show that the separation location is modified due to a decrease in the Reynolds shear stress within the boundary layer. This decrease in turbulent mixing, brought about by the presence of a superhydrophobic coating, reduces the transport of momentum to the wall which leads to earlier onset of turbulent boundary layer separation. This identifies a concern that should be addressed as superhydrophobic coatings are investigated as a method for reducing ship resistance within the maritime industry.