The aim of this research is to study the formation mechanism of Barrier Layers (BL) in the western tropical Atlantic and their influence on the tropical Atlantic climate at both short and long timescales. Many Coupled General Circulation Models (CGCMs) tend to overestimate the salinity in the Atlantic warm pool or the Northwestern Tropical Atlantic (NWTA) and underestimate the surface salinity in the subtropical salinity maxima region. Most of these models also suffer from a seasurface temperature (SST) bias in the NWTA region, leading to suggestions that the upper ocean salinity stratification may need to be improved in order to improve the BL simulations and thus the SST through BL-SST-Intertropical Convergence Zone (ITCZ) feedbacks. We used a CGCM to perform a set of idealized numerical experiments to understand the sensitivity of the BL and consequently SST in the NWTA region to freshwater flux and hence the upper ocean salinity strati cation. We find that the BL of the western tropical Atlantic is quite sensitive to upper ocean salinity changes in the Amazon River discharge region and the subtropical salinity maxima region. The BL phenomenon is further manifested by the formation of winter temperature inversions in our model simulations. However, in the region of improved BL simulation, the SST response is not statistically significant. SST response to Tropical Cyclones (TCs) is studied for the Atlantic region using a high-resolution coupled regional climate model (CRCM) and observational data sets. The presence of a BL, defined as the layer below the mixed layer that separates the base of the isothermal layer from the base of the isohaline layer, is found to modulate the SST response. The amplitude of TC-induced surface cooling is reduced by more than 35 percent in the presence of a BL, as a consequence of the weak thermal stratification. Furthermore, in locations when the BL exhibits a temperature inversion, TC-induced mixing can result in weak surface warming. BLs considerably reduce the rightward bias for tropical storms, but the effect is less conspicuous for TCs. The enthalpy flux into the atmosphere at the air-sea interface is enhanced by 16 percent and the increase in upper ocean potential energy due to TC-induced mixing is reduced by 25 percent in the presence of BLs. The results from the coupled model are supported by an observational analysis performed using re-analysis data sets, as well as data from Argo floats and TRMM satellite. As previous modeling and observational studies have indicated that the surface cooling caused by TC-induced mixing acts as a negative feedback for its intensity, results from our study suggest that BLs may have potential implications for TC intensity prediction.

Proefschrift--Amsterdam. / "Stellingen" (3 p.) inserted. "Litteratuur-opgave": p. [177]-187.

The broad aim of the work presented in this thesis is to investigate the wake of an airfoil under the combined effects of streamwise curvature and pressure gradient. This was accomplished by an experimental investigation using hot-wire anemometry and large eddy simulation (LES). The wake was generated by placing a NACA 0012 airfoil in a uniform stream of air, which is then subjected to an abrupt 90o curvature created by a duct bend. The experimental work was conducted in a subsonic open-return type wind tunnel. The test section measured 457 mm × 457 mm in cross-section and consisted of a 90o bend with radius-to-height ratio of 1.17. The symmetrical airfoil was of chord length (c) 150 mm, and its trailing edge was located one chord length upstream of the bend entry. The effects of airfoil angle of attack and mainstream velocity on the mean velocity and turbulence quantities of the near-wake were examined. In addition, the mean velocity and turbulence intensity profiles of the boundary layer on the upper surface of the airfoil were measured. In the numerical investigation, the three-dimensional, incompressible turbulent flow in the duct was computed using the finite volume method. The effect of modelling parameters, namely, grid resolution and sub-grid scale (SGS) model were studied. Three different sub-grid scale models were employed, namely, the classical Smagorinsky, its dynamic variant (DSMG) and the dynamic kinetic energy transport. The effect of grid resolution was assessed by conducting simulations with the DSMG model on three different grids. The first two grids incorporated the full spanwise extent of the duct (3c), and the third grid comprised a reduced spanwise segment (0.5c) with periodic conditions set in the spanwise direction. A bounded central differencing scheme was employed for the discretization of the convection terms. The temporal discretization was by a second-order implicit method that incorporated a forward difference approximation. The performance of LES in depicting the experimental flow was assessed and compared with the results predicted by the Reynolds Stress Model. The experimental profiles at zero angle of attack revealed the differing effects of curvature on the mean and turbulence quantities in the inner-side and outer-side of the wake. The spanwise distributions of mean velocity and turbulence intensity, in the near-wake, indicated variations with identifiable peaks and troughs which corresponded to the presence of streamwise vortices in the wake. The spanwise variations were larger on the inner side of the wake and significantly reduced on the outer side. The results showed that close to the trailing edge, the dominant effect on the wake was from the airfoil boundary layer, whereas one chord length downstream of the trailing edge, it was the effect of curvature and pressure gradient from the duct which was dominant. The results from the numerical study showed the advantages of LES over Reynolds-averaged Navier-Stokes methods in predicting separation on the convex wall of the bend on relatively coarse grids, but also shortcomings in the prediction of the wake parameters. The dynamic variants of the SGS models were more accurate in predicting the flow in the wake. On a considerably finer grid with near-wall airfoil grid spacings of Δx+ < 80, Δy+ < 0.5, and 20 < Δz+ < 50, LES resulted in much improved comparisons with the experimental data. The improved prediction of the wake parameters was attributed to the improved simulation of the boundary layers on the upper surface of the airfoil. However, the effect of the reduced spanwise extent resulted in a lack of prediction of separation on the convex wall of the duct.

The goal of the present work is a detailed and comprehensive study to assess the accuracy of the numerical simulation of the mixing processes in a lobed mixer flow field via a Reynolds-averaged solution method. To meet this goal, the first objective of the current work was to establish the suitability of various meshing strategies that would allow the complex mixer geometries found in current gas-turbine engine designs to be captured, together with the associated convoluted shear layers. A second objective was targeted at providing further insight and understanding of the capability of eddy-viscosity-based turbulence models in capturing the convoluted shear layers. Simplified mixer configurations selected from the literature were studied under incompressible isothermal flow conditions. Two solution algorithms were employed to model the mixer flow fields. The first consisted of a pressure-based structured grid methodology developed for incompressible flows. A density-based mixed-unstructured grid algorithm for compressible flows was also used, with extensions to low Mach number flows made possible through a low Mach number preconditioner. The effects of turbulence were modelled using ak-e turbulence model. The absence of this model in the code made available for the unstructured algorithm necessitated its implementation as a first step in the current work. The effects of unstructured mesh type on the prediction of flows with internal mixing layers were first assessed for an incompressible planar mixing layer. This simplified case was used as a benchmark case to help understand the effects on the convoluted shear layers arising within the lobed mixer flows. To quantify the capability of a Reynolds-averaged approach in simulating the turbulent mixer flow field, two variants of the two equation k-e model were employed. The first constituted the standard linear high Reynolds number k-e model of Launder and Spalding [62]. The second model was a quadratic non-linear version developed by Speziale [99] for the prediction of secondary flows in non-circular ducts. The relative merits of these two models was assessed through detailed comparisons with experimental data taken from the literature. Of particular importance in the mixer flow was the formation and subsequent evolution of the vorticity field. Consequently, this motivated a detailed study of the evolving vorticity field. The investigations thus far were based on a simplified mixer configuration with no temperature differences between the two streams. Therefore, as a final step, a realistic scarfed mixer was modelled in an attempt to model the temperature mixing. The main contribution of the present work is the assessment of a grid-based Reynolds-averaged solution procedure for the prediction of lobed mixer flows. The study revealed that capturing the initial mixing region proved to be most difficult. Firstly, unstruc-tured meshes employing non-hexahedral elements were very inefficient at simulating the mixing layer in the early stages. Secondly, the initial mixing region presented significant difficulties for the Reynolds-averaged solution method in which neither turbulence model was capable of correctly reproducing the turbulence field. Despite this, global parameters such as momentum thickness and streamwise circulation were well captured in the predictions.

Proefschrift--Amsterdam. / "Stellingen" (3 p.) inserted. "Litteratuur-opgave": p. [177]-187.

The oxidation behaviour of TiAlN based hard coatings with the addition of Cr and/or Y was investigated using scanning electron microscopy, scanning/transmission electron microscopy, energy dispersive X-ray analysis, thermogravimetry and X-ray diffraction. The coatings were deposited using the combined cathodic arc/unbalanced magnetron deposition technique. The main practical application of these films is dry high speed cutting in difficult to cut materials such as AISI A2 steel. Especially in the case of TiAlCrYN coating with an oxynitride and Cr-enriched overcoat, extensive research on the oxidation behaviour was performed and described here. Heat treatments in air between 600-1000°C for different duration were carried out. The Ti[0.44]Al[0.54]Cr[0.02]N coating was used as the starting point for the investigations. The effect of heat on the composition of the interface region was investigated. This region is of utmost importance for the adhesion of the film. In the case of TiAICrN the interface stability was not guaranteed because of diffusion of the substrate elements Cr and Fe to the coating surface after annealing for 1h at 900&deg;C. In comparison, the diffusion of substrate elements Cr and Fe in a ~2.3 mum thick coating of Ti[0.43]Al[0.52]Cr[0.03]Y[0.02]N and of Ti[0.34]Al[0.62]Cr[0.03]Y[0.01]N with overcoat, reached only a distance of ~600nm into the coating. This was achieved by the diffusion of Y to the grain boundaries. Y probably reacted at the same time with inward diffusing O. The diffusion of Y to the boundaries was observed after heat treatment for 1h at 900&deg;C or 10h at 800&deg;C.Ti[0.26]Al[0.26]N/Cr[0.48]N was the coating with the least oxide layer growth after oxidation for 1h at 900&deg;C. An oxide layer thickness of only ~100nm was measured. For the TiAICrYN coating with overcoat an oxide layer of 230nm and for TiAICrYN of 430nm formed after 1h at 900&deg;C. TiAlCrN in comparison formed an oxide layer of ~800nm after 1 h at 900&deg;C.The oxide layers formed after 1h at 900&deg;C consisted mainly of an Al[2]O[3] and TiO[2] bi-layer in the case of TiAlCrN and TiAlCrYN. The addition of a Cr-rich oxynitride overcoat led to the formation of a mixture out of Al[2]O[3], Cr[2]O[3] and TiO[2] in the oxide layer. In the case of TiAlN/CrN, a solid solution consisting of Cr[2]O[3] and Al[2]O[3] was observed. In general a stress relief after heat treatment was observed. At the same time the formation of voids along the column boundaries was identified. This was explained with the relaxation and diffusion of defects created during the deposition process. The effect of different substrate materials on the oxidation behaviour was also investigated. It was found that the formation of substrate oxides on the coating surface is very much dependent on the onset point of oxidation of the substrate material itself. The oxidation of substrate material occurred mainly through growth defects and pinholes. In cases where cracks formed during heat treatment of the coating, the formation of oxides out of substrate elements were observed in cracks connecting the substrate with the coating surface. Changing the bias voltage altered the formation of cracks. This research emphasises the importance of Y in the oxidation mechanism of TiAlN based hard coatings. Y blocks the diffusion path along the column boundaries and thus stowed down the diffusion and oxidation process. At the same time the addition of Cr can increase the oxidation resistance considerably, which was observed in the TiAlCrYN coating with and without overcoat.

This project is an elementary school located in a suburban area outside of Williamsburg, Virginia. The building is sited on an eight acre plot of wooded land and is designed for 100 students. A playground is designed as a counterpart to the school. A design approach which resembles the layering of space is developed through the drawings. The experience of these spaces establishes a dialogue between building and student likened to a game of hide and seek. / Master of Architecture

This thesis presents numerical investigations of two-dimensional single-phase turbulent jets and bubbly mixing layers using Reynolds-Averaged Navier-Stokes (RANS) equations. The behavior of a turbulent jet confined in a channel depends on the Reynolds number and geometry of the channel which is given by the expansion ratio (channel width to jet thickness) and offset ratio (eccentricity of the jet entrance). Steady solutions to the RANS equations for a two-dimensional turbulent jet injected in the middle of a channel have been obtained. When no entrainment from the channel base is allowed, the flow is asymmetric for a wide range of expansion ratio at high Reynolds number. The jet attaches to one of the channel side walls. The attachment length increases linearly with the channel width for fixed value of Reynolds number. The attachment length is also found to be independent of the (turbulent) jet Reynolds number for fixed expansion ratio. By simulating half of the channel and imposing symmetry, we can construct a steady symmetric solution to the RANS equations. This implies that there are possibly two solutions to the steady RANS equations, one is symmetric but unstable, and the other solution is asymmetric (the jet attaches to one of the side walls) but stable. A symmetric solution is also obtained if entrainment from jet exit plane is permitted. Fearn et al. (Journal of Fluid Mechanics, vol. 121, 1990) studied the laminar problem, and showed that the flow asymmetry of a symmetric expansion arises at a symmetry-breaking bifurcation as the jet Reynolds number is increased from zero. In the present study the Reynolds number is high and the jet is turbulent. Therefore, a symmetry-breaking bifurcation parameter might be the level of entrainment or expansion ratio. The two-dimensional turbulent bubbly mixing layer, which is a multiphase problem, is investigated using RANS based models. Available experimental data show that the spreading rate of turbulent bubbly mixing layers is greater than that of the corresponding single phase flow. The presence of bubbles also increases the turbulence level. The global structure of the flow proved to be sensitive to the void fraction. The present RANS simulations predict this behavior, but different turbulence models give different spreading rates. There is a significant difference in turbulence kinetic energy between numerical predictions and experimental data. The models tested include 𝘬—𝜖, shear-stress transport (SST), and Reynolds stress transport (SSG) models. All tested turbulence models under predict the spreading rate of the bubbly mixing layer, even though they accurately predict the spreading rate for single phase flow. The best predictions are obtained by using SST model. / Master of Science

<p>The purpose of this thesis was to study the influence of manufacturing variations on gear performance. The manufacturing variations inherent in different manufacturing methods were studied to include the effect of real surfaces. Real surfaces have surface irregularities at least on some scale, which can significantly influence how loads are transmitted at the gear contact. To some extent, the lubricant design can help to prevent contact that could lead to tooth failures by forming a protective surface boundary layer. An experimental study was used to consider the compositions of these layers with a surface analysis method.</p><p>In Paper <strong>A</strong> a robust design approach was used to find out to what extent the current standard for calculation of surface durability treats manufacturing variations and the choice of lubricant. The results show that the simplest calculation method used is not enough to predict the effect of these on surface durability. Additionally, the standard quality levels are poorly incorporated in the standard calculating procedures for surface durability, and the quality of the gear tooth is restricted to include only a few parameters.</p><p>In Paper <strong>B</strong> a pin-on-disc machine was used to evaluate the tribofilm formation by the additives and the corresponding wear occurring in the boundary lubrication regime in environmentally adapted lubricants. Studies of the additive and base fluid interaction were carried out using glow discharge-optical emission spectroscopy. It was found that the chemically reacted surface boundary layers played an important role in terms of wear. More specifically, the oxide layer thickness had significant influence on wear. The findings also demonstrate the complexity of lubrication design formulations coupled to these layers. For example, it was found that the pre-existing surface boundary layer (before any lubricant had been added) played an important role in allowing the lubricant to react properly with the surfaces.</p><p>The aim of Paper <strong>C</strong> was to contribute to the knowledge of how different surface topographies, tied to manufacturing methods, influence the early life contact conditions in gears. Topographical measurements of differently manufactured tooth flanks were used as data input to a contact analysis program. The variation in surface topography inherent in the manufacturing method was found to have a strong influence on the contact area ratio.</p>

No description available.