Geri Dön

A framework for composite wind turbine blades: From material development to structural optimisation and ml-enhanced damage detection

Rüzgar türbinleri için hibrit kompozit temelli kanat malzemesi geliştirilmesi, yapısal optimizasyonu ve makine öğrenmesi ile hasar tespiti ve yapısal analiz entegrasyonu

  1. Tez No: 1009525
  2. Yazar: KEMAL HASIRCI
  3. Danışmanlar: DOÇ. DR. ALAEDDİN BURAK İREZ
  4. Tez Türü: Doktora
  5. Konular: Makine Mühendisliği, Enerji, Mechanical Engineering, Energy
  6. Anahtar Kelimeler: Hibrit kompozit, Makine mühendisliği, Yapısal optimizasyon, Yenilenebilir enerji, Hybrid composite, Mechanical engineering, Structural optimizasyon, Renewable energy
  7. Yıl: 2026
  8. Dil: İngilizce
  9. Üniversite: İstanbul Teknik Üniversitesi
  10. Enstitü: Lisansüstü Eğitim Enstitüsü
  11. Ana Bilim Dalı: Makine Mühendisliği Ana Bilim Dalı
  12. Bilim Dalı: Makine Mühendisliği Bilim Dalı
  13. Sayfa Sayısı: Belirtilmemiş.

Özet

Rüzgâr enerjisi, sürdürülebilir enerji üretimine geçişte küresel ölçekte en önemli stratejik kaynaklardan biri hâline gelmiştir. Temiz enerjiye olan talebin dünya genelinde hızla artmasıyla birlikte, rüzgâr türbinleri sera gazı emisyonlarının azaltılmasında ve uluslararası karbon nötr hedeflerine ulaşılmasında kilit bir rol oynamaktadır. Rüzgâr türbinlerinin en kritik ve aerodinamik açıdan en hassas bileşenlerinden biri olan kanatlar, hem üretim aşamasında hem de operasyonel bakım maliyetlerinde aslan payını oluşturmaktadır. Bu bileşenler, karmaşık geometrileri ve çalışma prensipleri gereği yüksek mekanik yükler, dinamik kuvvetler ve değişken çevresel etkilere maruz kalmaları nedeniyle mühendislik tasarımı açısından büyük zorluklar barındırmaktadır. Özellikle deniz üstü (offshore) türbin kanatları, karasal sistemlere kıyasla çok daha şiddetli rüzgâr rejimleri, sürekli tuzlu su korozyonu ve tekrarlı yorulma yükleri gibi aşırı koşullar altında işlev görmektedir. Bu zorlu çalışma ortamı, malzemenin mikro yapısında zamanla geri döndürülemez hasarların oluşmasına neden olmakta; bu da daha dayanıklı, hafif, kendi hatasını sönümleyebilen ve akıllı denetim mekanizmalarına sahip yeni nesil kanat sistemlerinin geliştirilmesini zorunlu kılmaktadır. Bu doktora tezi,“Rüzgâr Türbini Kanatlarında Malzeme Geliştirme, Yapısal Optimizasyon ve Makine Öğrenmesi ile Denetim”başlığı altında, malzeme bilimi, yapısal mekanik ve yapay zekâ disiplinlerini bir araya getiren üç ana araştırma alanını bütünleştirmektedir. Çalışma kapsamında; (1) mikro hasarların ilerlemesini durdurarak hasar toleransını artırmaya yönelik kendi kendini onarabilen (self-healing) gelişmiş kompozit malzemelerin sentezlenmesi, (2) hibrit kompozit kanat geometrilerinin ağırlık ve maliyet dengesini kurmak amacıyla makine öğrenmesi destekli vekil modeller üzerinden yapısal optimizasyonu, ve (3) insansız hava araçlarından alınan verilerin derin öğrenme ve fizik tabanlı modelleme ile birleştirildiği akıllı kusur tespit ve mekanik değerlendirme sistemleri ele alınmıştır. Bu üçlü sacayağı, rüzgâr türbini kanatlarının sadece tasarım aşamasını değil, aynı zamanda operasyonel ömrünü ve bakım süreçlerini kapsayan bütüncül bir dijital ikiz ve akıllı izleme çerçevesi ortaya koymaktadır.

Özet (Çeviri)

Wind energy has become a cornerstone of the global transition toward sustainable energy production. As the demand for clean power intensifies, wind turbines play an increasingly vital role in reducing greenhouse gas emissions and supporting decarbonization initiatives. Among their key components, the blades are the most structurally demanding and technologically sophisticated parts, accounting for a significant portion of both manufacturing and maintenance costs. Offshore wind turbines, in particular, are exposed to extreme mechanical stresses and environmental degradation due to high winds, saltwater corrosion, and continuous fatigue loading. Consequently, the development of durable, lightweight, and intelligent blade systems is essential for improving turbine performance, safety, and lifecycle economics. This doctoral dissertation titled“Wind Turbine Blade Material Development, Structural Optimization, and Inspection Using Machine Learning”aims to address these challenges by integrating three interrelated research domains: (1) the development of advanced self-healing composite materials for enhanced damage tolerance, (2) machine learning–driven structural optimization for lightweight and high-strength hybrid blades, and (3) intelligent inspection systems utilizing deep learning and physics-based modeling for automated defect detection and evaluation. Together, these efforts establish a holistic framework for the design, operation, and maintenance of next-generation smart wind turbine blades. Chapter 1 provides an overview of the thesis objectives, research hypotheses, and a comprehensive literature review. The primary purpose of the study is to enhance the mechanical robustness, reliability, and sustainability of wind turbine blades through the synergy of materials engineering, computational modeling, and artificial intelligence. The literature survey examines recent advances in blade materials, including glass and carbon fiber composites, hybrid laminates, and self-healing systems. It also reviews the limitations of conventional design and inspection techniques, such as high computational cost in optimization processes and the dependence on manual visual inspections for damage assessment. The chapter formulates the central research hypothesis that integrating self-healing materials, data- driven optimization, and machine learning–based inspection can significantly extend the service life and efficiency of wind turbine blades while reducing maintenance costs. The chapter concludes by outlining the methodological flow: experimental material development, numerical optimization, and data-centric inspection analysis. Chapter 2 focuses on experimental research dedicated to the development of self- healing and tribologically enhanced laminate composites for wind turbine blades. The first study, Development of Self-Healing Glass Fiber Reinforced Laminate Composites for Wind Turbine Blades, explores the incorporation of microencapsulated healing agents into glass fiber reinforced polymer (GFRP) laminates using the vacuum-assisted resin transfer molding (VARTM) method. Mechanical performance was characterized through tensile and Charpy impact tests, supported by microscopic evaluations to assess crack propagation and healing behavior. The findings indicate that a 2.5% concentration of self-healing microcapsules offers the most favorable balance between strength and healing efficiency. Additionally, increasing the curing temperature from 80°C to 100°C enhances the ultimate tensile strength (UTS), demonstrating the sensitivity of healing performance to thermal curing conditions. The second study, Tribologically Enhanced Self-Healing Hybrid Laminates for Wind Turbine Applications, extends this concept by introducing silicon carbide whiskers (SiCw) as reinforcing fillers to counteract potential reductions in mechanical properties due to microcapsule addition. The hybrid composite system was subjected to tensile, impact, and tribological testing. Results revealed a 32% increase in UTS and a 45% improvement in energy absorption with the addition of SiCw. Furthermore, the hybrid laminate exhibited a 10% increase in wear resistance and a 20% decrease in the friction coefficient, demonstrating superior tribological performance. After an adequate healing period following impact, the system achieved a 5% recovery in tensile strength. This study establishes that combining self-healing mechanisms with tribological reinforcements can yield multifunctional composites with improved mechanical and surface durability, suitable for next-generation wind turbine blades. Chapter 3 presents a two-stage Machine Learning–Driven Surrogate Modeling Approach for the Structural and Cost Optimization of Hybrid Composite Wind Turbine Blades. The study integrates finite element analysis (FEA) with machine learning to achieve efficient, accurate, and economically viable optimization of large- scale composite structures. Using a 61.5-meter, 5 MW reference blade, a parametric finite element model was constructed in Abaqus to simulate different stacking sequences of unidirectional glass, unidirectional carbon, and twill carbon laminates. Mechanical inputs were derived from experimental data obtained in Chapter 2 to ensure model fidelity. The initial optimization objectives focused on minimizing blade weight and tip displacement simultaneously while satisfying the Tsai–Wu failure criterion, utilizing trained surrogate models to accelerate the process. This first stage achieved a 15.6% reduction in blade weight with negligible loss of stiffness. To further enhance economic feasibility, a second-stage, section-based optimization was introduced, which selectively applied hybrid laminates to high-load regions and glass fiber composites to less critical zones. This refinement reduced the total blade cost by 19% while maintaining safety standards. The resulting framework offers a scalable, computationally efficient tool for developing wind turbine blades that are both structurally optimized and cost-effective. Chapter 4, A Hybrid Framework for Damage Assessment in Composite Wind Turbine Blades via Deep Learning and Numerical Modeling, introduces a novel inspection methodology that integrates UAV-based imaging, deep learning–driven defect detection, and finite element–based structural evaluation. The system enables real-time monitoring and assessment of offshore wind turbine blades, overcoming the limitations of conventional manual inspections. UAV-acquired images are analyzed using a YOLOv11 deep learning model, achieving a mean average precision (mAP@50) of 95.6% with 95.4% precision and 95.8% recall. Detected defects such as surface cracks, erosion, adhesive failures, and impact-induced delaminations are converted into geometric representations for FEA in Abaqus. The corresponding simulations evaluate stress distributions, fracture mechanics parameters, and damage propagation trends. This integrated deep learning–FEA pipeline establishes a direct link between visual detection and mechanical criticality assessment, enabling predictive maintenance and informed decision-making. The approach represents a significant advancement in autonomous structural health monitoring (SHM) for offshore wind turbines. The final chapter summarizes the contributions of the thesis and discusses future research directions. The results demonstrate that integrating self-healing composite systems, AI-driven optimization, and intelligent inspection provides a comprehensive pathway toward high-performance, self-aware wind turbine blades. The findings not only contribute to the advancement of material science and computational mechanics but also offer practical implications for reducing maintenance costs, improving reliability, and extending the operational life of wind turbines. Collectively, the outcomes of this research lay the foundation for smart, sustainable, and efficient wind energy technologies that align with the goals of a carbon-neutral future.

Benzer Tezler

  1. Tez No

    863672

    Capacity analysis in Turkey for carbon fiber production considering offshore wind turbines demand forecast

    Açık deniz rüzgar türbinleri talep tahminine göre Türkiye'de karbon elyaf üretimi için kapasite analizi

    SEZGİN KARABACAK

    Yüksek Lisans

    İngilizce

    İngilizce

    2024

    Endüstri ve Endüstri Mühendisliğiİstanbul Teknik Üniversitesi

    İşletme Mühendisliği Ana Bilim Dalı

    PROF. DR. HÜR BERSAM SİDAL

  2. Tez No

    797260

    Development of progressive failure models and their application to open hole composite laminates and wind turbine blades

    İlerlemeli hasar modellerinin geliştirilmesi ve delikli kompozit plakalar ve rüzgar türbin kanatlarına uygulanması

    MEVLÜT CAN MUYAN

    Doktora

    İngilizce

    İngilizce

    2023

    Havacılık ve Uzay MühendisliğiOrta Doğu Teknik Üniversitesi

    Havacılık ve Uzay Mühendisliği Ana Bilim Dalı

    PROF. DR. DEMİRKAN ÇÖKER

  3. Tez No

    762956

    El yatırma yöntemi ile lif takviyeli kompozit üreten bir işyerinde iş sağlığı ve güvenliği risklerinin değerlendirilmesi

    Manufacturing fiber reinforced composites by hand lay-up method assessment of occupational health and safety risks

    YASİN AYTEPE

    Yüksek Lisans

    Türkçe

    Türkçe

    2022

    Mühendislik BilimleriBursa Teknik Üniversitesi

    Biyokompozit Mühendisliği Ana Bilim Dalı

    PROF. DR. SAMİ İMAMOĞLU

  4. Tez No

    66441

    Sabancı Center Akbank Kulesi için bir sistem analizi

    The System analysis for Akbank Tower of Sabancı Center

    OSMAN SALICI

    Yüksek Lisans

    Türkçe

    Türkçe

    1997

    İnşaat Mühendisliğiİstanbul Teknik Üniversitesi

    Yapı Ana Bilim Dalı

    PROF. DR. ERDOĞAN UZGİDER

  5. Tez No

    504561

    Cephe kaplama sistemlerinin performans gereksinimleri açısından değerlendirilmesi

    Evaluation of facade cladding systems with regard to performance requirements

    BÜŞRA SİRKECİ

    Yüksek Lisans

    Türkçe

    Türkçe

    2017

    Mimarlıkİstanbul Teknik Üniversitesi

    Mimarlık Ana Bilim Dalı

    PROF. DR. LEYLA TANAÇAN

Geri Dön