Istanbul Technical University
Ulusal Üniversitelerarası Açık Erişim Sistemi - İstanbul Teknik ÜniversitesiNot a member yet
67356 research outputs found
Sort by
Elmas içerisinde renk merkezleri oluşturularak kuantum emiterlerin üretilmesi
No full textThesis (M.Sc.) -- Istanbul Technical University, Graduate School, 2025This thesis explores the generation of nitrogen-vacancy (NV) color centers in diamond, which are solid-state quantum emitters. Quantum emitters are critical for applications in quantum cryptography, sensing, and computing. Due to their stable photoluminescence and long spin coherence times at room temperature, NV centers are particularly worthy of studying in quantum technologies. The study aims to produce NV centers in a controlled manner by using low-energy electron beam irradiation and thermal annealing. The experimental method began with the selection of a nitrogen doped high-pressure high-temperature (HPHT) synthesized type Ib single-crystal plate diamond with a {100} surface orientation, obtained from Element Six. Electron irradiation was carried out using a lithography system at 100 keV with varying doses (0.5–5.0 C/cm²) applied in specific patterns on four corners of this sample. This step aimed to introduce vacancies in a spatially resolved manner. Subsequent annealing at 800 °C under argon flow at 100 CCM was done to induce the migration of these vacancies to nearby substitutional nitrogen atoms. These processes combined lead to the formation of NV centers. A comprehensive characterization process was conducted at each stage using topographical imaging with an optical profilometer, Raman spectroscopy, photoluminescence (PL) spectroscopy, and electron spin resonance (ESR). Raman analysis confirmed the integrity of the diamond lattice post-treatment. It also provided insights into strain–stress relations within the diamond lattice. Shifts in the diamond Raman peak position and changes in full width at half maximum (FWHM) across the sample revealed localized lattice stress. This discovery was important because these stress variations directly influence NV center formation and performance. PL spectroscopy detected distinct zero-phonon line (ZPL) emissions at ~638 nm, indicative of negatively charged NV⁻ centers. The intensity of NV⁻ emission was observed to vary with electron dose, which indicates the ability to control NV center density. Mapping of the PL signal further demonstrated successful spatial control of NVs on the sample. ESR measurements confirmed the spin-related properties of the generated centers and the charge state transition from NV⁰ to NV⁻ following annealing. Overall, this work stands out for its use of low-energy electron irradiation which is preferable due to easier control over defect positioning and potentially lower damage to the host lattice. This thesis contributes to the development of deterministic NV center engineering.M.Sc
Sağlık hizmetlerinde metaverse teknolojileri: bibliyometrik bir analiz, Tıp alanında kabul süreçlerinin ve Tıp fakültelerinde ders çizelgeleme stratejilerinin incelenmesi
No full textThesis (Ph.D.) -- Istanbul Technical University, Graduate School, 2025The metaverse, once a concept confined to science fiction, has evolved into a technological paradigm that integrates virtual, augmented, mixed, and extended reality. These immersive tools have increasingly entered healthcare and medical education, enabling new approaches to clinical training, remote therapy, surgical simulation, and collaborative learning. The COVID-19 pandemic accelerated this trend by disrupting traditional education and highlighting the need for scalable and digitally accessible environments. Nevertheless, the academic literature on the healthcare metaverse remains fragmented and methodologically limited. This thesis addresses these gaps by investigating the adoption and implementation of metaverse technologies in healthcare, with a particular focus on medical education. A multi layered methodological approach is adopted, combining bibliometric mapping, behavioral modeling, and mathematical optimization to provide both theoretical and practical insights. Chapter 2 presents a bibliometric analysis of 3,721 publications indexed in the Web of Science between 1994 and 2022. The studies are categorized into thematic groups, and productivity patterns are assessed through Zipf's, Lotka's, and Bradford's laws. The findings highlight influential authors, countries, institutions, and underexplored themes, establishing a solid foundation for subsequent analyses. Chapter 3 examines the behavioral intentions of medical doctors in Turkey using the Technology Acceptance Model extended with technology anxiety. Structural Equation Modeling validates the framework and reveals that perceived usefulness, ease of use, and anxiety significantly affect adoption. Chapter 4 expands the model to include both practitioners and medical students. By integrating complementary theories such as Innovation Diffusion Theory, the Interaction Equivalence Theorem, and Embodied Social Presence Theory, the study demonstrates that satisfaction, presence, interaction, trialability, compatibility, readiness, and imagination all play important roles in acceptance. Chapter 5 develops a novel course scheduling optimization model to integrate metaverse based courses into medical curricula. The model incorporates professor preferences derived from Structural Equation Modeling and prioritized with the Analytic Hierarchy Process. A binary integer programming formulation is solved using Greedy and Simulated Annealing algorithms, showing that the latter achieves more balanced workloads and higher instructor satisfaction. The thesis concludes by emphasizing its theoretical contribution by broadening acceptance models with psychological and sociotechnical factors and its practical contribution by offering a replicable framework for institutions modernizing curricula through immersive technologies. Limitations include reliance on synthetic optimization data and the need for real world validation. Future studies should focus on longitudinal, cross cultural, and adaptive approaches. Overall, the thesis provides an integrative perspective on the role of the metaverse in healthcare education by uniting bibliometric, behavioral, and optimization methods, thereby enriching the literature and offering actionable guidance for decision makers.Ph.D
Driven segmentation of synthetic microwave images for breast cancer detection
No full textTez (Yüksek Lisans)-- İstanbul Teknik Üniversitesi, Lisansüstü Eğitim Enstitüsü, 2025Breast cancer remains one of the most prevalent and life-threatening diseases worldwide, continuing to be the leading cause of cancer-related mortality among women. Current global health statistics reveal that millions of new breast cancer cases are diagnosed annually, with a disproportionate concentration in low- and middle-income countries (LMICs). In these regions, disparities in healthcare infrastructure, socioeconomic challenges, and limited access to advanced diagnostic technologies substantially hinder early detection efforts. Early diagnosis of breast cancer is of paramount importance, as timely intervention significantly improves patient survival rates, reduces the need for aggressive treatments, mitigates treatment-related side effects, and ultimately decreases the economic burden on healthcare systems. Despite technological advancements, conventional breast imaging modalities such as mammography, ultrasound, and magnetic resonance imaging (MRI) still face notable limitations globally in terms of accessibility, cost, and diagnostic efficacy. Mammography, regarded as the clinical gold standard for breast cancer screening, involves ionizing radiation and poses potential health risks with frequent use. Moreover, its diagnostic sensitivity is considerably reduced in women with dense breast tissue, leading to false-negative results and delayed treatment. Radiation-free ultrasound imaging requires highly skilled operators, resulting in variability in diagnostic accuracy and limiting its widespread application. Nuclear imaging techniques such as positron emission tomography (PET) and single-photon emission computed tomography (SPECT) provide valuable dynamic and functional imaging data; however, their high costs, radiation exposure, and limited availability restrict their broader clinical use. MRI offers superior soft tissue contrast and spatial resolution but is associated with high operational costs, infrastructure demands, and limited accessibility, especially in LMICs. Collectively, these challenges underscore the urgent need for safe, cost-effective, scalable, and reliable novel imaging technologies for early tumor detection. Microwave Imaging (MWI) has recently emerged as a promising alternative modality that enables non-ionizing electromagnetic wave-based examination of the internal structural and functional properties of breast tissue. The fundamental principle of MWI relies on the dielectric property differences—specifically dielectric permittivity (∈_r) and electrical conductivity (σ)—between malignant tumors and normal breast tissue. This contrast provides functional imaging information complementary to the anatomical data obtained from conventional methods. Microwave imaging systems primarily consist of transmitting and receiving antennas, with the collected data computationally reconstructed into images. Additionally, MWI systems are inherently more affordable, portable, and easier to operate, making them highly suitable for resource-constrained clinical environments. Nevertheless, the clinical translation of MWI has been impeded by intrinsic limitations such as relatively low spatial resolution, sensitivity to noise and artifacts, and challenges in accurately localizing tumors within reconstructed images. In response to these challenges, the present thesis proposes a comprehensive deep learning framework aimed at enhancing the diagnostic performance and tumor localization accuracy of microwave breast imaging. Within this scope, synthetic breast tissue models were developed using a high-fidelity simulation platform based on the Reverse Time Migration (RTM) algorithm, which realistically models the propagation and scattering of electromagnetic waves in heterogeneous and homogeneous biological tissues. The RTM algorithm processes source and receiver wavefields bidirectionally and establishes correlations between them to generate an image. It consists of two steps involving the backpropagation of complex conjugate data into the imaging domain and computing the cross-correlation norm of the forward and backward propagated fields. Synthetic breast models were generated employing microwave difference imaging based on the total electric field, necessitating knowledge of the healthy breast state during imaging. The total electric field data required for microwave imaging were synthetically produced via numerical solutions of electromagnetic scattering problems. This study investigates deep learning-based segmentation of low-resolution microwave images generated from total electric field data. The dataset comprises 1000 synthetic transverse breast images, each containing a single tumor, initially created at a high resolution of 400×400 pixels. The use of synthetic data instead of physical breast phantoms provided precise control over critical parameters such as tissue heterogeneity, tumor size and location, and dielectric properties, allowing systematic and reproducible evaluation of model robustness under diverse conditions. The imaging configuration simulates a circular array of 18 Vivaldi antennas placed around the synthetic breast at a radius of 10 cm. The surrounding medium is modeled as a homogeneous environment with a relative permittivity of 10. Simulations incorporate properties of breast, fibroglandular, and tumor tissues. This arrangement ensures wide angular coverage and enhances sensitivity to tissue heterogeneity. To align with deep learning architectures and reduce computational load, images were resized to 64×64 pixels while preserving essential structural and contrast information necessary for accurate segmentation. It should be noted that microwave images are not inherently RGB; rather, single-channel microwave data were artificially converted into three-channel (64, 64, 3) pseudo-color images using assigned color maps. This transformation enables universal convolutional neural networks designed for RGB inputs to process the data. Both deep learning models used in this thesis accept input and produce output images of (64, 64, 3) dimensions, visually highlighting tumor regions to facilitate interpretation and improve suitability for segmentation tasks. Within this framework, two distinct deep learning architectures were proposed for tumor segmentation: an autoencoder and a U-Net model. Both architectures were implemented in Python using libraries such as Keras, TensorFlow, skimage, matplotlib, and numpy. Both models focus primarily on semantic segmentation of tumor tissues beyond mere noise removal or artifact suppression. The dataset of 1,000 images was split into 600 for training, 200 for validation, and 200 for testing on unseen samples. Training and testing were conducted in two separate phases; first, the systems were trained and the architectures saved, followed by testing using the trained models. The autoencoder comprises an encoder that compresses input images through multi-layer convolutional blocks incorporating batch normalization and LeakyReLU activation functions. Batch normalization accelerates training and stabilizes intermediate outputs, while LeakyReLU prevents "dead ReLU" problems by allowing small non-zero gradients for inactive neurons. Dropout regularization in deeper layers prevents overfitting. The encoder compresses fundamental spatial and structural features into latent representations, which the decoder reconstructs into enhanced images emphasizing tumor regions through transposed convolutions and additional convolutional operations. This architecture not only reconstructs image content but also improves contrast and structural integrity to support subsequent segmentation steps. However, since the autoencoder primarily aims at general image enhancement, its performance in precise tumor localization remains limited. In contrast, the U-Net model employed in this study is a supervised architecture specifically designed for biomedical image segmentation. U-Net features a symmetric encoder-decoder structure interconnected by skip connections that preserve high-resolution spatial information losslessly. The encoder extracts multi-scale features via convolution and max-pooling layers, while the decoder progressively upsamples these features to generate tumor probability maps. A sigmoid activation function in the final output layer produces binary segmentation masks with well-defined tumor boundaries. Unlike the autoencoder, U-Net is directly optimized for semantic segmentation using custom loss functions that consider both spatial accuracy and tumor region integrity. Consequently, it yields highly precise and clinically meaningful segmentation maps suitable for decision-making. Both models maintain structural symmetry between compression and expansion rates to ensure performance and stability during training. They were trained and validated on different subsets of the synthetic dataset with extensive data augmentation including rotation, reflection, and scaling to improve generalization and prevent overfitting. The Adam optimizer was used during training, with the autoencoder trained for 150 epochs and the U-Net for 50 epochs. Structural Similarity Index (SSIM) was employed as the primary evaluation metric to assess preservation of structural details and perceptual image quality during both training and testing phases. Comparative results demonstrate that the autoencoder effectively enhances overall image contrast and suppresses irrelevant details but remains limited in tumor localization due to its reconstruction-centric design. Conversely, the U-Net model significantly outperforms the autoencoder in both SSIM scores and segmentation accuracy, producing highly reliable tumor segmentation masks. In conclusion, this thesis presents a deep learning-based framework addressing fundamental limitations of microwave breast imaging. Utilizing high-fidelity synthetic datasets and robust neural network architectures, the proposed approach substantially improves tumor detection and localization capabilities. The non-invasive, radiation-free, and low-cost nature of microwave imaging renders this technology particularly promising for expanding early breast cancer screening in resource-limited settings worldwide. The proposed methods close current technological gaps by enabling more accurate tumor detection in low-resolution microwave images. Moreover, the successful integration of microwave imaging with deep learning not only overcomes existing technical challenges but also paves the way for more accessible and equitable breast cancer diagnostic methods globally. This approach reduces dependence on expensive and complex imaging equipment, holding transformative potential for screening protocols in disadvantaged regions where early detection remains critically insufficient. The innovations presented in this thesis represent a significant step toward reducing global health disparities, providing clinicians with advanced tools for timely diagnosis and personalized treatment planning. Ultimately, this work offers meaningful progress in the fight against breast cancer by combining technological innovation with intelligent healthcare solutions to enhance patient survival rates and quality of life.Günümüzde yapay zekâ tabanlı derin öğrenme modelleri, görüntü işleme ve bilgisayarlı görü alanlarında önemli gelişmelere öncülük etmektedir. Bu gelişmeler, özellikle tıbbi görüntüleme alanında hastalıkların erken teşhisi ve doğru tanılanması açısından büyük fayda sağlamaktadır. Kadınlarda en yaygın şekilde karşılaşılan kanser türlerinden biri meme kanseridir ve erken teşhis, tedavi başarısını doğrudan etkilemektedir. Bu nedenle, güvenli, taşınabilir ve iyonlaştırıcı radyasyon içermeyen bir yöntem olan mikrodalga görüntüleme, alternatif bir görüntüleme tekniği olarak dikkat çekmektedir. Ancak mikrodalga görüntülerin düşük çözünürlüklü ve gürültülü olması, doğrudan yorumlanmalarını güçleştirmektedir. Bu çalışmada, mikrodalga görüntülerde tümörlü dokuların daha net ve doğru şekilde belirlenebilmesi amacıyla derin öğrenme tabanlı bir segmentasyon yöntemi kullanılmıştır. Bu sayede, düşük kaliteli görüntülerden daha anlamlı bilgiler elde edilerek meme kanseri teşhisinde yardımcı olunması hedeflenmiştir. Bu çalışmada, bir adet tümör dokusu içeren sentetik meme yapısına ait toplam 1000 adet mikrodalga görüntü veri seti ve bu veri setine karşılık gelen segmentasyon görüntüleri girdi verisi olarak kullanılmıştır. Görüntülerin oluşturulabilmesi için 10 cm yarıçaplı 18 adet vivaldi anten içeren özel bir senaryo tasarlanmıştır. Bu senaryoda antenler, sentetik meme dokusunun etrafına yerleştirilerek farklı açılardan sinyal gönderip alacak şekilde konumlandırılmıştır. Bu sayede, tümörlü bölgenin mikrodalga sinyallerine verdiği tepkiler kaydedilmiş ve bu veriler kullanılarak dielektrik görüntüler oluşturulmuştur. Bu dielektrik görüntüler Reverse Time Migration (RTM) algoritması kullanılarak mikrodalga görüntülere çevrilmiştir. Elde edilen bu görüntüler, derin öğrenme tabanlı modellerin eğitimi ve testinde kullanılmıştır. Bu çalışmada iki farklı mimari kullanılmıştır: Autoencoder ve U-Net. Bu mimariler Python yazılım dili kullanılarak tasarlanmıştır. Bu iki mimari için üretilen 1000 adet görüntünün 600'ü eğitim, 200'ü validasyon ve 200'ü test için ayrılmıştır. Autoencoder modeli, gürültüyü azaltmak amacıyla girdiyi kodlayıp yeniden oluşturan bir yapıya sahiptir. Optimizasyon algoritması olarak Adam optimizasyonu kullanılmıştır. U-Net mimarisi ise gürültüyü temizlerken detayları daha iyi koruyabilmek amacıyla geliştirilmiştir. Eğitim sürecinde Binary Crossentropy kayıp fonksiyonu kullanılmıştır. Mimarilerin başarımı, Yapısal Benzerlik İndeksi (SSIM) metriği ile değerlendirilmiştir. SSIM değeri 0 ile 1 aralığındadır ve sonucun 1 olması orijinal görüntü ile tahmin edilen görüntünün birbiriyle tamamen aynı olduğu anlamına gelmektedir. Yapılan çalışma sonucunda autoencoder mimarisinin SSIM değeri 0,9419 iken, U-Net mimarisinin SSIM değeri 0,9627'dir. Bu sonuçlar doğrultusunda U-Net Mimarisi, autoencoder mimarisine göre %2,21 oranında mikrodalga görüntü segmentasyonunda daha başarılı olmuştur.M.Sc
Determination of shear lag in high-rise reinforced concrete systems and its effects on analysis and design
No full textTez (Yüksek Lisans)-- İstanbul Teknik Üniversitesi, Lisansüstü Eğitim Enstitüsü, 2025Tüp tipi taşıyıcı sistemler, yüksek dayanımlı malzemelerin verimli kullanımı ve yanal yük taşıma kapasiteleri sayesinde çağdaş yüksek katlı yapı tasarımlarında giderek daha fazla tercih edilen sistemler hâline gelmiştir. Ancak bu sistemler, kesit boyunca eksenel yüklerin üniform dağılmaması ile karakterize edilen "kayma gecikmesi (shear lag)" olarak adlandırılan bir etkiye maruz kalmaktadır. Kayma gecikmesi, özellikle tüp sistemlerin yanal yük taşıma verimliliğini düşüren temel bir problem olup, bu etkinin yapı sistem parametreleriyle olan ilişkisi henüz yeterince detaylı şekilde incelenmemiştir. Bu çalışma, kayma gecikmesi davranışını etkileyen üç temel yapısal sistem değişkenini; yanal rijitlik, kat planı en-boy oranı ve yapı yüksekliği araştırmayı amaçlamaktadır. Bu amaçla, SAP2000 yapısal analiz programı kullanılarak altı adet sonlu eleman modeli oluşturulmuştur. İlk olarak, yanal rijitliğin kayma gecikmesine etkisini değerlendirebilmek adına, farklı sayıda kuşak kirişi içeren üç model geliştirilmiştir: kuşak kirişi içermeyen model A; bir adet kuşak kirişi içeren model B; iki adet kuşak kirişi içeren model C ve üç adet kuşak kirişi içeren model D. Yapılan karşılaştırmalar sonucunda, kuşak kirişi sayısı arttıkça yanal rijitliğin arttığı ve buna bağlı olarak kayma gecikmesi oranlarının azaldığı görülmüştür. Ayrıca kayma gecikmesinin pozitif değerlere ulaştığı kat seviyesinin, yanal rijitlik arttıkça üst katlara doğru ötelenme eğiliminde olduğu tespit edilmiştir. İkinci olarak, kat planı en-boy oranının etkilerini incelemek amacıyla, model A ile aynı kat yüksekliğine ve sistem tipine sahip, ancak x yönünde açıklığı iki kat artırılmış model E oluşturulmuştur. Bu sayede plan düzleminde x/y oranı değiştirilmiş, yapının y yönü sabit tutulmuştur. Model A ile model E arasında yapılan karşılaştırma sonucunda, yöne dik kenar (flanş) üzerindeki kayma gecikmesi oranlarının yaklaşık iki kat arttığı; yöne paralel kenarde (web) ise bu oranların büyük ölçüde sabit kaldığı görülmüştür. Son olarak, yapı yüksekliği değişkeni incelenmiştir. Başlangıçta oluşturulan model F, 35 katlı bir yapı olarak tasarlanmış ve 30 katlı model A ile karşılaştırılmıştır. Elde edilen ilk bulgular, yapı yüksekliği arttıkça kayma gecikmesi oranlarının azaldığını göstermiştir. Bu sonuç, mevcut literatürde genellikle kabul gören "yüksekliğin kayma gecikmesini artırdığı" yönündeki görüşlerle çelişmektedir. Bu durumun geçerliliğini test etmek amacıyla, ilave olarak 32, 40 ve 50 katlı modeller oluşturularak analizler genişletilmiştir. Tüm bu modellerde benzer bir eğilim gözlenmiş ve kat sayısı arttıkça kayma gecikmesi oranlarının düştüğü, eksenel yük dağılımının daha üniform hale geldiği belirlenmiştir. Bu durumun olası sebebi olarak, yapı yüksekliği arttıkça eğilme etkilerinin baskın hale gelerek kayma gecikmesi etkisini gölgelemesi ve kolonlar arası eksenel yük dağılımını dengelemesi düşünülmektedir. Sonuç olarak bu tez kapsamında, üç farklı yapısal sistem değişkeninin kayma gecikmesi üzerindeki etkisi detaylı biçimde analiz edilmiş; tüp sistemlerin performansını etkileyen temel faktörler bütüncül olarak değerlendirilmiştir. Elde edilen sonuçlar, mevcut literatüre katkı sağlayacak nitelikte olup, özellikle yapı yüksekliği gibi bazı parametrelerin etkilerine dair alternatif yorumlar geliştirilmesine olanak sunmuştur. Bu çalışma, hem tasarım pratiği açısından yol gösterici olmayı hem de ileride yapılacak benzer araştırmalara altyapı oluşturmayı hedeflemektedir.Tube-type structural systems have emerged as a widely adopted solution in the design of contemporary high-rise buildings due to their high efficiency in resisting lateral loads and their economical use of high-strength materials. These systems, often formed by closely spaced perimeter columns connected by deep spandrel beams, provide a stiff and strong exterior "tube" that resists lateral forces induced by wind or earthquakes. Their ability to significantly reduce the structural demand on interior framing and core walls has made them especially favorable for tall building applications. However, a critical challenge associated with tube systems is the occurrence of a phenomenon known as shear lag, which significantly impacts the effectiveness of axial force transfer along the building's height. Shear lag refers to the non-uniform distribution of axial forces across the cross-section of a structure. This phenomenon results in certain edge columns carrying higher or lower loads than anticipated under ideal uniform stress assumptions, thereby diminishing the structural efficiency. In particular, this uneven axial load distribution is more pronounced under lateral loading conditions where bending dominates. Despite the increasing prevalence of tube systems in practice, the influence of key structural parameters on shear lag behavior remains inadequately understood. Thus, this thesis aims to systematically investigate the impact of three fundamental system parameter which are lateral stiffness, plan aspect ratio, and building height on the shear lag behavior in tube-type structural systems. To address this objective, a parametric study was conducted using CSI SAP2000 version 18, a widely used finite element analysis software. A total of six analytical models were developed and subjected to lateral loading scenarios simulating seismic action. The modeling approach adhered to typical assumptions in tall building design: rigid diaphragm behavior at floor levels, fixed-base support conditions, and moment-resisting connections at beam-column joints. All models employed equivalent structural configurations, varying only in the parameter being evaluated, ensuring a consistent basis for comparison. The results were interpreted by comparing axial force distributions, calculating shear lag ratios, and observing the transition of stress patterns across floor levels. The first parameter examined was lateral stiffness, assessed by altering the number of belt trusses which is horizontal bracing systems that provide additional stiffness and structural integrity to framed tube systems. Four distinct models were created to capture the variation in stiffness: Model A, which had no belt truss; Model B, with a single belt truss at mid-height; Model C, incorporating two belt trusses placed at quarter and three-quarter heights; and Model D, equipped with three evenly spaced belt trusses. The comparative analysis of these models demonstrated a direct relationship between increased lateral stiffness and reduction in shear lag effects. As more belt trusses were added, the lateral rigidity of the system improved, resulting in more balanced axial force distributions among columns. Additionally, the story level at which the shear lag ratio became positive was observed to shift higher in taller models, indicating delayed onset of stress concentration at the flanges as stiffness improved. The second parameter investigated was the plan aspect ratio, defined as the ratio of horizontal dimensions in orthogonal directions. Model E was created by modifying Model A to have double the span in the x-direction, thus increasing the x/y plan ratio while maintaining all other geometric and structural properties constant. The goal was to isolate the effect of plan geometry on shear lag behavior. Analysis of Model E in comparison to Model A revealed a considerable increase in shear lag values in the flange regions which is the sides perpendicular to the seismic load direction. In these regions, the axial forces concentrated more at the corners, while the web sides, parallel to the seismic load, showed negligible changes. This result aligns with the expectation that increasing the width of the structure intensifies the bending moments in the transverse direction, exacerbating the differential force distribution among flange columns and thus amplifying the shear lag effect. The final parameter examined was building height, a factor frequently cited in literature as contributing to increased shear lag due to amplified flexibility and global deformations. To explore this, Model F was designed as a 35-story counterpart to the original 30-story Model A. Contrary to expectations, Model F exhibited lower shear lag ratios. This finding prompted further validation, leading to the development of three additional models with 32, 40, and 50 stories. Each of these models showed a similar pattern: shear lag ratios decreased as building height increased. The load distribution became more uniform, particularly in upper floors. This trend indicates that with taller structures, bending behavior begins to dominate, producing a smoother axial stress profile across the structure. The increasing curvature and flexibility in tall buildings contribute to an equalization of stress among perimeter columns, thereby reducing shear lag intensity. This discovery contradicts prevailing assumptions in the field and suggests that height, in isolation, does not exacerbate shear lag but interacts with other structural parameters to influence overall behavior. In summary, the findings from this study reveal that all three structural system parameters which are lateral stiffness, plan aspect ratio, and building height play a significant role in the formation and mitigation of shear lag. The inclusion of belt trusses effectively enhances lateral stiffness and reduces axial stress differentials. An increased plan aspect ratio tends to worsen shear lag effects in flange regions due to higher bending moments. Surprisingly, increased building height contributes to a more balanced axial force distribution, possibly due to bending-induced stress harmonization. These insights challenge conventional design wisdom and offer a more nuanced understanding of shear lag phenomena in high-rise buildings. This research has important implications for both academic and practical fields. For practicing engineers, the study offers quantifiable evidence on how specific design decisions influence shear lag performance. Optimizing belt truss placement and reconsidering plan proportions could lead to more efficient use of materials and safer structural performance under lateral loads. For researchers, the thesis opens new questions about the nonlinear interactions between system parameters and their impact on stress distribution, particularly in very tall or irregularly shaped buildings. Future studies may incorporate material nonlinearity, time-dependent effects such as creep and shrinkage, or dynamic behavior under real earthquake records to further enrich the understanding of shear lag mechanisms. Furthermore, this thesis encourages a reconsideration of current building codes and design practices, which often address global stability and strength requirements but fall short of providing specific guidance for mitigating shear lag. Integrating findings like those in this study into design manuals and structural guidelines would contribute to more informed, performance-based design approaches in high-rise construction. As cities continue to grow vertically, understanding the subtleties of load distribution in complex structural systems becomes increasingly crucial. In conclusion, this study provides a comprehensive analysis of how shear lag is influenced by key structural parameters in tube-type high-rise buildings. The findings not only clarify the relationships among stiffness, geometry, and height, but also provoke a deeper dialogue about the assumptions embedded in tall building design. By addressing these relationships through a parametric and computational lens, the research contributes to more accurate, resilient, and efficient structural systems in the field of high-rise architecture and engineering.Yüksek Lisan
Artçı depremlerin çelik yapılardaki etkisi
No full textThesis (M.Sc.) -- Istanbul Technical University, Graduate School, 2025This thesis provides an extensive investigation into the seismic performance of steel moment-resisting frames that undergo mainshock-aftershock (MS-AS) sequences, while also considering the impact of several aftershocks and retrofitting options. Nonlinear time history analysis was invoked to assess 3-, 9-, and 20-story SAC buildings modelled and analyzed with real ground motion records that represent either near-fault or far-fault ground shaking. The intended novelty was the realistic modelling of sequential earthquake events (including extended aftershocks) to be able to quantify their cumulative structural damage effects. The analysis began by assessing the original (non-retrofitted) frames subjected to either mainshock-only and MS-AS sequences. The most important observation was that aftershocks could influence interstory drift ratios and base shear demands. This was especially significant in the short and mid-rise structures. Damage accumulation is considered the evolution and distribution of plastic hinges. Of the three frames, the 3-story frame experienced the most significant increase in terms of drift and hinge damage due to aftershocks, while the 20-story building was the most stable relative to overall behavior in the early stages of the MS-AS sequences. Seismic resilience can be improved through retrofitting methods that rely on concentric braces in exterior bays of moment-resisting frames. The retrofitted frames showed a considerable improvement in performance. While it was anticipated that base shear would increase to be consistent with the additional lateral stiffness found in the retrofitted frames, the maximum interstory drifts were considerably decreased. The data demonstrates that retrofitting to achieve seismic resilience will significantly reduce damage and improve the overall structural performance under repeated seismic loading and aftershocks. In the extended special case, the 3-, 9-, and 20-story buildings were subjected to up to five consecutive aftershocks following a mainshock, with the sequences derived from the Mammoth Lakes earthquake. The analysis revealed that structural responses (i.e. maximum drift and base shear) vastly increased after the first and second aftershocks, however, structural response plateaued after the third aftershock, with only marginal additional drift or damage observed with subsequent events. This plateau indicates an upper bound for the damage state for the structure, or redistribution of internal forces within the system, which likely precluded further damage or drift. Notably, while the number of plastic hinges appeared to increase for the 20-story model, the number of CP-level hinges did not increase proportionally, which suggests that tall buildings are tolerant when subjected to sequential earthquakes. These results signify the importance of including aftershock sequences and cumulative seismic demand in the seismic design and assessment of structures. Even if designers only consider the loads from a mainshock on their designs, it is easy to suspect that they potentially underestimate the actual risk of seismic loading, and this study has shown that seismic load in the presence of aftershocks for previously designed and built structures is far from negligible. Additionally, this study shows that retrofitting involves planning simple designs and that retrofitting enhances deformation and damage control, ensuring safety during multiple earthquake events. To conclude, this study contributes to understanding the realistic structural behavior of steel moment-resisting frames subjected to sequential earthquakes. The inclusion of the aftershock effect in the design and retrofitting of steel buildings is essential to achieve a safer and resilient structure.M.Sc
TDMK 2025
No full text10. Türkiye Deprem Mühendisliği Konferansı, TMMOB İnşaat Mühendisleri Odası İstanbul Şubesi ile Türkiye Deprem Vakfı – Deprem Mühendisliği Komitesi ortaklığında 8–9–10 Ekim 2025 tarihlerinde İstanbul Teknik Üniversitesi Süleyman Demirel Kültür Merkezi’nde gerçekleştirilecektir. Bu yıl ikinci kez uluslararası nitelikte düzenlenen konferansımız, 1985 yılında gerçekleştirilen ilk sempozyumdan bu yana ülkemizde ve dünyada deprem mühendisliği alanındaki güncel bilgi birikimini, deneyimleri ve yaklaşımları bir araya getiren önemli bir bilimsel platformdur.TMMOB İnşaat Mühendisleri OdasıTürkiye Deprem Vakf
Rasyonel kübik splayn fonksiyonu kullanarak iki ve üç boyutlu veri yapılarının analizi ve görselleştirilmesi: İstanbul'daki su, doğal gaz ve elektrik verileri üzerine bir vaka çalışması
No full textThesis (M.Sc.) -- Istanbul Technical University, Graduate School, 2025Spline is an interpolation technique that we use to best connect a series of data points in a mathematical sense and obtain a curved function. Basically, they consist of functions between every two data points, which we call multi-part polynomials, which are used to connect certain data points with a smooth curve. The function derivatives here are continuous. This means that the curve does not create corner breaks by making a smooth transition. Spline is also used to provide smooth transitions between data points, to create continuous curves without sharp corners or sudden changes, and to prevent excessive oscillations, i.e. unnecessary fluctuations, in higher-degree polynomial interpolations. Because in splines, instead of higher-degree polynomials, piecewise functions are used. It is preferred to use it when working with continuous data sets in data visualization, modeling data trends, engineering mathematics, computer graphics, numerical analysis, climate science, bioinformatics, machine learning and artificial intelligence. When we apply spline interpolation to real data sets, we obtain functions that provide smooth and continuous transitions between the data points we have. They are also used to estimate missing or noisy data, to determine trends or to model the dynamics of a system. There are many different spline methods in the literature. Different data sets and applications require the use of different spline types. The reason for choosing different spline types is the accuracy, calculation and shape control requirements that vary depending on the structure of the data set and modeling needs.In this thesis, different spline types are examined and linear spline, cubic spline and rational cubic spline methods known in the literature are used. While linear spline provides a fast and simple solution, cubic spline increases interpolation accuracy with smooth transitions. Rational cubic spline is advantageous in providing additional control conditions such as positivity. In this study, the rational cubic spline method has been examined in detail. There are many different forms of rational cubic spline. In this study, we use the form with four free parameters and C^1 continuity conditions. Shape control analyses have been performed to stretch the curve bends of these parameters or to provide control on the curve. The ability of shape control analyses to provide the desired shapes of parametric curves has been examined. In particular, in the analyses performed on rational cubic spline curves, it has been examined how well these curves fit the function. Rational cubic spline is a method that can preserve the positivity of the curve with the help of shape control parameters. Positivity is important in data that has a physical meaning, such as the data sets used here, such as water, natural gas and electricity consumption, which cannot be negative. In addition, three different derivative selection methods were used for rational cubic spline: Arithmetic method, geometric method and central difference method. Different derivative selections were evaluated in order to increase the accuracy of the curve and modeling success. While the arithmetic method provides an average transition, the geometric method captures proportional changes better. The central difference method increases the accuracy of interpolation with the symmetric derivative calculation approach. Peano Kernel Theorem is used to determine how close the rational cubic spline curve is to the function, to test the accuracy of spline derivatives, and to determine the upper bounds of the error in differential and integral error estimates. Peano Kernel Theorem is a method used in error analysis of spline interpolation and is used to compare and analyze the approximation errors of different spline types. As a result of these analyses, it is revealed which parameters should be optimized in what way in order to increase the precision of spline interpolation. The real datasets we used in this study are water, natural gas and electricity consumption data of Istanbul province. The datasets were obtained from the official website of Istanbul Metropolitan Municipality (IMM) and Istanbul Water and Sewerage Administration (ISKI). Python programming language was used in the visualizations. It is necessary to preserve positivity in rational cubic spline interpolation for the water, natural gas and electricity consumption datasets used in the study. Because water, natural gas and electricity consumptions can never be negative. They gain meaning when they are positive. For this reason, it is necessary to use a method that does not allow negative values and preserves positivity in the interpolation process. The spline interpolation methods used here, especially the rational cubic spline, were used to ensure that the consumption amounts in the dataset remain compatible with physical reality. Linear spline, cubic spline and rational cubic spline interpolation methods applied to real data sets provide a smooth and understandable representation of water, natural gas and electricity consumption data in time series format. Thanks to these methods, values that are not measured or are made but missing can be estimated. Precise analyses can be performed on consumption data. Since water, natural gas and electricity consumption quantities do not show a regular change over time, spline interpolations are sensitive to changes and fluctuations that may occur and can provide results accordingly. The use of these interpolation methods and visualization techniques provides valuable information for municipalities, energy and water distribution companies, environmental researchers and city planners. With these visualizations and models, the examination of consumption trends is very important in terms of efficient use of resources, establishing supply-demand balance and sustainability. In such cases, taking precautions against any possible situations is important in planning maintenance and repair processes and planning infrastructure investments. In two-dimensional visualizations, linear spline, cubic spline and rational cubic spline were used and derivatives for rational cubic spline were calculated with arithmetic, geometric and central difference methods. Each interpolation and derivative calculation method is colored so that users can compare the performance of different methods. Two-dimensional visualizations present the change of consumption data over time in a simple and understandable way, while allowing for clearer observation of increases or decreases in a certain period. In addition, error analyses were performed to evaluate the accuracy of interpolation methods and compare their performances. Five different interpolation metrics were used to perform these error analyses. These metrics were determined as mean absolute error MAE, mean square error MSE, coefficient of determination R^2, mean square logarithmic error RMSE and sum of squared error SSE. In this way, the accuracy and reliability levels of different methods on the data sets could be compared. Bicubic spline and rational bicubic spline interpolation methods were used for three-dimensional visualizations. In order to obtain three-dimensional images, the SciPy package in the Python programming language was used for cubic spline interpolation. For rational bicubic spline images, a rectangular region was defined and the rational cubic spline function was expanded. Through these visuals, water and natural gas consumption of 39 different districts of Istanbul province was presented in detail with changes on a yearly and monthly basis. Three-dimensional visualizations allow for a more holistic examination of temporal and spatial consumption patterns, allowing for more comprehensive analyses for both decision makers and researchers. Working with data sets and comparing various applications also helps evaluate consumption habits of different geographical, climatic and demographic characteristics. As a result, the use of such interpolation and visualization approaches supports faster, strategic, technical and effective decision-making in areas such as energy and water management, and contributes to the development of various policies for a more resilient infrastructure and resource management against possible problems in the future.M.Sc
Insights into Fatigue Damage in Additive Manufacturing through Process-Structure-Property Interactions
No full texthttps://doi.org/10.1007/s11668-025-02305-
An energy approach to pulsar–disc interaction: disc stability and implications for transitional millisecond pulsars
No full textABSTRACT The stability of an accretion disc surrounding a millisecond pulsar is analysed from an energetic point of view, using magnetohydrodynamic simulations that consider realistic disc structures and a variety of magnetic field inclination angles. The time-averaged components of the magnetic field interact with the disc through ohmic dissipation, which causes heating and partial evaporation of its innermost region. The stability of the disc right after the magnetic field is turned on is analysed as a function of the location of the inner radius of the disc and the magnetic inclination angle. Our results show that the disc is severely altered in those cases where its inner radius lies well beyond the light cylinder and the magnetic axis is not totally aligned with the neutron star spin axis. Overall, the results of the simulations agree with those obtained in previous works where analytical or semi-analytical energy models were also used to discuss the stability of the disc. The implications for the understanding of the transitional millisecond pulsars are discussed. We briefly mention implications of our results for low-mass X-ray binaries and supernova fallback discs.https://doi.org/10.1093/mnras/staf854https://dx.doi.org/10.48550/arxiv.2505.23407http://arxiv.org/abs/2505.23407http://hdl.handle.net/10261/399152https://api.elsevier.com/content/abstract/scopus_id/105009376496http://dx.doi.org/10.13039/501100000781http://dx.doi.org/10.13039/501100002809http://dx.doi.org/10.13039/501100011033http://dx.doi.org/10.13039/501100000780http://dx.doi.org/10.13039/501100004351http://dx.doi.org/10.13039/501100003329http://dx.doi.org/10.13039/50110000483
Investigating the performance of super-resolved remote sensing images on coastline segmentation with deep learning based methods
No full textThe use of satellite imagery in critical areas, such as environmental monitoring and natural disaster management, is becoming increasingly important. Applications like monitoring coastal areas, detecting coastal erosion, and tracking land use changes demand high accuracy and detailed analysis. Traditional methods for coastline segmentation are often limited by the low resolution (LR) and high complexity of satellite imagery. To address this challenge, Super Resolution (SR) algorithms are employed to enhance the resolution of satellite images, which is particularly beneficial when examining areas with intricate structures, such as coastlines. In this context, the integration of SR and segmentation techniques presents an innovative approach to achieving greater accuracy and efficiency in satellite image analysis. In this study, the resolution of satellite images was enhanced using the Super Resolution Generative Adversarial Networks (SRGAN) model. Thanks to the flexible architecture of the SRGAN model, it was successfully adapted to work with satellite images, yielding satisfactory results. Coastal segmentation was performed using low-resolution, super-resolved, and high-resolution Gokturk-1 (GT-1) satellite images, employing U-net, LinkNet, and DeepLabV3+ segmentation models for comparison. The results indicated that increment in image resolution significantly affects segmentation success. Additionally, better performance in coastline segmentation was achieved with U-net and LinkNet models. Although the DeepLabV3+ model is effective for segmentation, it tends to capture less detail compared to the other two models. Overall, the combination of SRGAN and the LinkNet segmentation model produced results that were closest to realityhttps://doi.org/10.26833/ijeg.1522143https://dergipark.org.tr/tr/pub/ijeg/issue/90124/152214