5 publications — 2 first-author journals (under review) · 2 co-authored (1 under review, 1 in preparation) · 1 conference paper
MSc Thesis
📖
M.Sc. Thesis · Manufacturing Engineering · Sabancı University · July 2026
Integrating Multi-Sensor Experimental and Numerical Approaches for Damage Localization and Crack Propagation Monitoring in Composite Laminates
İbrahim Bahadır
· Supervisor: Assoc. Prof. Adnan Kefal
This thesis presents an integrated methodology for damage characterization and crack propagation monitoring in pre-cracked carbon/epoxy composite laminates. To address the numerical sensitivity of full-field optical measurements, a dual-stage smoothing element analysis (SEA) assisted digital image correlation (DIC) methodology is developed to formulate a gradient-enhanced, bounded damage-sensitive localization index. This numerical formulation identifies spatially localized damage zones without relying on constitutive material models. Utilizing a unified experimental campaign on seven composite laminates with varying stacking sequences (0/90/0, 90/0/90, 0/0/0, 90/90/90) and initial crack orientations (45º and 90º), the crack propagation behavior is comprehensively investigated through a multi-instrument approach combining DIC, acoustic emission (AE), and strain gauges. Results demonstrate that the macroscopic mechanical response and microscopic damage evolution are strongly governed by both ply architecture and initial crack angle. Specifically, 0º-dominated laminates exhibit progressive horizontal crack propagation, whereas 90º-dominated laminates display delayed but abrupt vertical fracture. AE clustering analysis identifies two distinct acoustic event populations corresponding to tensile crack-tip activity and high-energy mixed-mode sliding mechanisms. By synthesizing the spatial deformation from DIC, temporal damage activity from AE, and local confirmation from strain gauges, this thesis provides a robust framework for tracking both progressive crack growth and abrupt high-energy fracture events.
Digital Image CorrelationAcoustic EmissionSmoothing Element AnalysisComposite LaminatesCrack PropagationDamage Localization
Full-field crack propagation monitoring in composite laminates using digital image correlation and smoothing element analysis
İbrahim Bahadır,
M. H. Bilgin,
A. Al-Nadhari,
M. Yildiz,
A. Kefal
Under Review
Elsevier · Composite Structures
Crack propagation and crack-sensitive localization in composite laminates are difficult to identify reliably from experimental full-field measurements because DIC-derived strain fields are sensitive to noise amplification during differentiation and further instability during spatial-gradient evaluation. This study proposes a dual-stage smoothing element analysis (SEA)-assisted digital image correlation (DIC) framework for full-field crack propagation monitoring in cracked composite laminates. The measured in-plane DIC displacement field is used as the primary input. In the first stage, SEA regularizes the displacement field before strain reconstruction, while in the second stage, the von Mises equivalent strain field is smoothed before gradient evaluation. A gradient-enhanced, thresholded, and nonlinearly normalized damage-sensitive localization index is formulated directly from experimental displacement data, without requiring a constitutive damage model, finite element damage simulation, or material degradation parameters. The framework is assessed using seven cracked laminate specimens with different stacking sequences and 45º or 90º initial crack orientations. The displacement verification shows that the SEA-reconstructed fields remain in close agreement with raw DIC measurements while providing smoother and more stable fields for strain reconstruction. The reconstructed strain fields also show consistency with strain-gauge measurements at strain gauge locations. The localization maps capture the evolution of crack-sensitive regions before final fracture, including progressive horizontal localization in 0º-dominated laminates and delayed vertical localization in 90º-dominated laminates. The quantitative pre-failure assessment further shows that the high-index regions form close to the initial crack tips and align with the experimentally observed crack-propagation directions.
Keywords: Digital image correlation; smoothing element analysis; crack propagation monitoring; composite laminates; damage localization; gradient-enhanced localization index
First AuthorInternational Journal of Damage Mechanics · SAGE2026
Integrated multi-instrument analysis for damage characterization and crack propagation monitoring in pre-cracked composite laminates
İbrahim Bahadır,
M. H. Bilgin,
A. Al-Nadhari,
M. Yildiz,
A. Kefal
Under Review
SAGE · Int. J. Damage Mech.
Crack propagation in composite laminates is difficult to interpret using a single measurement technique because the damage process involves spatially localized deformation, time-dependent acoustic activity, and local strain redistribution. This paper presents an integrated multi-instrument experimental framework for damage characterization and crack propagation monitoring in pre-cracked composite laminates under tensile loading. Digital image correlation is used to capture full-field surface strain localization around the crack region, acoustic emission is used to monitor the temporal evolution of damage-related activity, and strain gauges are used in selected configurations to provide local mechanical confirmation. The framework combines global stress-strain response, DIC-based equivalent strain fields, AE time-history indicators, cluster-wise cumulative energy evolution, and local strain measurements within a stage-wise interpretation strategy. The results show that crack propagation is strongly affected by both laminate stacking sequence and initial crack orientation. DIC reveals the development of crack-related localization paths, while AE identifies the onset, accumulation, and intensification of damage activity. The cluster-wise AE analysis further distinguishes characteristic acoustic event populations without assigning them directly to fixed damage modes. For the strain-gauged configurations, local strain responses support the interpretation of damage stages by showing slope changes, unloading, or abrupt strain variations during critical intervals. Overall, the combined DIC-AE-strain gauge interpretation provides a more consistent basis for distinguishing progressive crack growth, delayed high-energy propagation, and abrupt fracture-dominated transitions in pre-cracked composite laminates.
Co-Author (2nd)Structural Control and Health Monitoring · Wiley2026
Experimental Shape Sensing of a NACA-Profile Composite Wing with FBG Sensors on a Single Surface using Inverse Finite Element Method
E. Yıldırım,
İbrahim Bahadır,
J. Bardiani,
C. Sbarufatti,
A. Kefal
Under Review
A single sensor based inverse finite element method (iFEM) framework is experimentally implemented and validated for the real-time deformation reconstruction of a tapered composite aircraft wing with an airfoil cross-section. The approach utilizes an enhanced iFEM formulation capable of modeling the complex, non-prismatic geometry of an airfoil, which is a significant advancement over simplified beam models. A key advantage of this methodology is the use of a sparse instrumentation network with Fiber Bragg Grating (FBG) sensors located on only a single surface, which addresses a critical implementation barrier of traditional back-to-back sensor configurations. First, the structural design of a hollow NACA 64010 profile is performed to obtain experimentally measurable deflections and rigorously test the accuracy of the sparse single-sided iFEM formulation. Then, the high accuracy of its finite element model is successfully demonstrated by comparing its strain predictions to experimental measurements, thereby establishing it as a reference solution. Afterwards, the predictive capabilities of the iFEM framework are examined on the manufactured composite wing using the experimental FBG strain measurements for two different load scenarios, i.e., pure bending and coupled bending-torsion. The quantitative comparison reveals an exceptionally high level of accuracy with reconstructed maximum deflection errors averaging only 0.6% for the pure bending case and 2.0% for the complex bending-torsion case. Overall, the single-sided iFEM remains robust even across a large non-instrumented wing region, supporting its practical potential for next-generation aircraft structural monitoring.
Keywords: Aircraft wing; fiber-reinforced composite; numerical and experimental mechanics; inverse finite element method; shape sensing; structural health monitoring
Co-Author (2nd)Journal · To Be SubmittedIn Prep.
Experimental Validation of Particle Inverse Method for Full-Field Deformation Reconstruction and Crack Propagation Monitoring of Composites Using Digital Image Correlation Measurements
M. H. Bilgin,
İbrahim Bahadır,
A. Kefal
In Preparation
Conference Papers
Conference · Co-Author (2nd)
EWSHM 2026 · 12th European Workshop on Structural Health Monitoring2026
Experimental Validation of Particle Inverse Method for Shape Sensing of Damaged Composite Structures from Discrete Sensor Measurements
M. H. Bilgin,
İbrahim Bahadır,
A. Kefal
Presented
Reliable reconstruction of full-field displacements and monitoring of crack growth in composite structures from limited sensing are critical needs for in-service structural health monitoring. This study presents an experimentally validated workflow for strain-driven shape sensing and crack propagation assessment in a damaged unidirectional carbon-epoxy laminate using the Particle Inverse Method (PIM). A three-ply composite plate is manufactured with an initial central crack oriented perpendicular to the fibre direction. During a quasi-static tensile test, surface-mounted strain gauges record discrete strain measurements on one face, while Digital Image Correlation (DIC) measures full-field displacements on the opposite face. Although DIC provides accurate displacements, strain fields obtained by direct differentiation are typically noisy and become unreliable near displacement discontinuities introduced by cracking. To address this limitation, DIC displacement fields are extracted at selected time steps spanning pre- and post-propagation regimes and processed using Smoothing Element Analysis (SEA) to obtain continuous, physically consistent strain fields. The SEA-computed strains are quantitatively verified against strain-gauge readings at sensor locations, yielding a dense and validated strain dataset for the cracked specimen. These verified strain fields are then used as the sole input to PIM to reconstruct full-field displacements and infer crack evolution through the method’s damage-aware particle interactions, without requiring prior knowledge of material properties or applied load magnitudes. The reconstructed displacement fields exhibit strong agreement with DIC across all time steps, while the predicted crack propagation behaviour is consistent with the experimentally observed failure response. The results demonstrate a practical DIC-SEA-PIM pathway to convert experimentally accessible measurements into decision-ready full-field kinematics and crack monitoring capability for damaged composite structures.
For a complete and up-to-date list of my publications, including citation metrics: