Boğaziçi University
4.002
Arşivlenen Tez
0
DOI Atanmış
7
Enstitü
0%
DOI Oranı
Son Eklenen Tezler
Tümünü GörEffect of loading rate on fracture of NiTi
This thesis studies the effect of loading rate on fracture behavior of Shape Memory Alloy (SMA) NiTi. First, Differential Scanning Calorimetry (DSC) are used to obtain the phase transformation temperatures of NiTi. Then series of tensile tests are performed for determining the mechanical properties under different loading rates. Next, fracture tests under increasing loading rates are conducted using pre-cracked, Mode-I Compact Tension (CT) specimens and critical stress intensity factors are calculated using the empirical equations proposed by ASTM E399 assuming that Linear Elastic Fracture Mechanics (LEFM) is applicable. Then, displacement field obtained with the Digital Image Correlation (DIC) technique is used for stress intensity factor (SIF) calculations. Phase transformation region extension along the crack is estimated with DIC measurements and is compared to the analytical calculations. Analytical calculations for phase transformation region size are carried out using SIFs obtained from DIC and from the empirical equations of ASTM E399. Finally, using material properties obtained from tensile tests and the results of fracture tests, Finite Elements (FE) models are built to calculate the phase transformation regions. Using the J contour integrals, stress intensity factors are estimated using LEFM relations. Using the displacement field obtained from the FE models, SIF values are calculated and the Crack Mouth Opening Displacement (CMOD) values are obtained from the FE models.
Active wave attenuation by a submerged heaving and pitching flat plate
Protecting the coastline and structures in its vicinity are main tasks of coastal engineering. Here an environmental-friendly and cost-efficient active submerged flat plate (SFP) breakwater is proposed. It is investigated whether through introduction, phase and amplitude control of two motion modes, heaving and pitching, and in-situ adaptations of submergence depth and plate inclination in dependence on the incidence wave, wave attenuation can be improved and the bandwidth of appreciable performance can be widened. It is examined whether maximum wave attenuation results in energy loss of the system so that the energy can be harvested and used for motion control and structural adaptations. The wave/structure boundary value problem (BVP) is formulated in two dimensions (vertical) and covers monochromatic waves only, linear wave theory is applied; nonlinearities are introduced by applying the body-nonlinear method. Motion amplitude and incidence wave height restrictions are enforced to ensure that the assumptions for linear wave theory are not violated. Plate motion is introduced through prescribed forcing with incident wave frequency. The BVP is solved numerically using the radial basis function (RBF) collocation method. A novel lateral boundary condition is developed that allows wave incidence and radiation simultaneous. The RBF method proved to be able to model wave/structure interaction and plate motion; model calibration is required. Through the applied body-nonlinear method higher harmonics are introduced. Research results suggest that wave attenuation of the SFP is improved through above-stated measures and performance bandwidth widened. Motion phase angle control is paramount to ensure destructive interference between scattered and radiated wave on the lee side of the plate. Configurations with minimum wave transmission allow for energy harvesting.
Improved thermal management of electrical vehicles usinginternal magnetic field of the electric motors
Improved cooling performance is required to ensure system durability and reliability of electric motors used for electric vehicles, and using magnetohydrodynamics principles offers a promising approach to enhance cooling performance. In this study, the utilization of the internal magnetic field of a 110 kW permanent magnet synchronous motor that has six axial cooling channels embedded between the coils to improve cooling performance is presented. The magnetic field for this motor is estimated and its effect on cooling performance is identified through numerical analysis. Magnetohydrodynamic effects are achieved using Cu-water/ethylene glycol magnetic nanofluids at varying particle volume fractions (up to 1%) and volume flow rates from 25 L/min to 35 L/min. The maximum improvement in thermal performance is observed at a flow rate of 25.8 L/min with a 1% volume fraction, yielding a remarkable 64% increase in the convection heat transfer coefficient and a 21% reduction in total thermal resistance, and 20% increase in cooling capacity by utilizing the internal magnetic field effect. The utilization of magnetohydrodynamic effects leads to a notable reduction in motor temperatures, thereby enhancing the endurance limits by up to 218%. The utilization of magnetohydrodynamics effects in this study proves advantageous to enhance the overall performance of the cooling system in the motor.
Multimodal statistical analysis of brain connectome with application to alzheimer's disease
Clinical cognitive decline, often culminating in Alzheimer's Disease Dementia (ADD), has traditionally been viewed as a disconnection syndrome affecting the brain's structural and functional organization. Despite extensive research focusing on changes in brain connectome models, particularly in the functional domain, a coherent understanding of underlying mechanisms and robust biomarkers remains elusive. Evidence suggests that the brain undergoes spatially non-stationary and non-monotonous structural and functional changes across the clinical decline continuum. In this thesis, these changes across different stages of clinical cognitive decline were examined in the context of local structural and functional reorganization in the brain. For this purpose, nodal assortativity was used as a quantitative metric. The findings reveal statistically significant local reorganization patterns in both structural and functional domains, which transition with the progression of cognitive decline and differ in nature between structure and function. Specifically, hypo-assortativity tends to prevail in the initial phases for both structural and functional aspects, transitioning to hyper-assortativity in later stages. Notably, functional reorganization is more pronounced in the early stages, while structural reorganization becomes prominent later. Despite some common regions that undergo both structural and functional reorganization, a significant variation was observed in regions affected from functional and structural perspectives. It is underscored that analyzing both spatial and temporal dimensions is crucial for a comprehensive understanding of clinical cognitive decline leading to ADD. The preliminary findings suggest a complex and variable reorganization process, potentially indicative of simultaneous active degenerative and compensatory mechanisms.
Life cycle assessment of energy and value-added product generation from biomass by pyrolysis
The sustainable management of biomass waste from food and raw materials is a critical issue today. Meanwhile, global energy demand is constantly increasing with the rapidly growing population and industrialization. Today, most of the energy used is still fossil fuels, and this dominance is a significant obstacle to the sustainable energy transition. The European Green Deal aims to make a sustainable economy possible by drawing attention to the transformation of climate and environmental problems brought about by these problems. At this point, pyrolysis technology offers a promising solution. It provides a cleaner way of waste management by providing valuable chemicals from biomass waste and energy generation. It enables us to use resources efficiently in line with the principles of circular economy. In this study, the environmental performance of a biomass waste pyrolysis plant, which obtains energy and valuable products by pyrolysis of biomass waste, was evaluated using the life cycle assessment methodology, and scenarios in which energy is assessed in different ways were discussed. In addition, changes in environmental impacts of this clean production plant, which contributes to the circular bioeconomy, were examined in the changing 2030 and 2053 electricity projections. The results show that environmental impacts are mainly due to the electricity consumption used in the production processes throughout the life cycle and the exhaust emissions of the cogeneration unit, and the system has the potential to make a positive contribution to abiotic depletion potential by producing activated carbon from biochar and converting biomass into energy.
Unveiling limitations in single-view 3D object reconstruction models
3D object reconstruction models typically learn from a single dataset, ShapeNetCore, and are evaluated against similar datasets that measure aspects closely related to ShapeNetCore. We tackled the problem of limited performance assessment by proposing novel benchmarks to reveal their robustness to new challenges. To demonstrate our benchmark's effectiveness, we selected three state-of-the-art models for comparison: 3D-C2FT, Pix2Vox, and Occupancy Networks. This selection covers two well-known 3D shape representations: voxel and occupancy function as an implicit representation. We first investigated the effect of changing background color on performance. We found that this seemingly simple variable causes a drastic decrease in performance. We observed that models perform sufficiently close to the original scenario with changing input object sizes in 2D. Further, we adapted a novel dataset 3DCoMPaT++ for 3D reconstruction evaluation. 3DCoMPaT++ offers rich material and part annotations. We assessed reconstruction performance by slightly changing viewpoints and varying styles in 2D input images. The results show that models struggle to adapt to novel settings. Performance degrades drastically in the best-case scenario, and surprisingly, the model performing the best in the standard ShapeNetCore experiment, Pix2Vox, scores the worst across all novel dataset experiments. We also evaluated models at the part level to identify the most challenging parts. We transferred part-level point clouds to part-annotated surface points from 3DCoMPaT++ using point cloud registration with L2 distance. We then utilized our version of F-Score@0.01, Part F-Score@0.01, for evaluation. This experiment quantitatively confirmed the known issue of poor performance in finer details and thin parts, unlike previous works that only made qualitative observations.