Olesya I. Zhupanska’s research while affiliated with University of Arizona and other places

In this work, a novel unsupervised machine learning (ML) method for automatic image segmentation of low velocity impact damage in carbon fiber reinforced polymer (CFRP) composites has been developed. The method relies on the use of non-parametric statistical models in conjunction with the so-called intensity-based segmentation, enabling one to determine the thresholds of image histograms and isolate the damage. Statistical distance metrics, including the Kullback–Leibler divergence, the Helling distance, and the Renyi divergence are used to formulate and solve optimization problems for finding the thresholds. The developed method enabled rigorous and rapid automatic image segmentation of the grayscale images from the micro computed tomography (micro-CT) scans of the impacted CFRP composites. Sensitivity of the segmentation results with respect to the thresholds obtained using different statistical distances has been investigated. Based on the analysis of the segmentation results, it is concluded that the Kullback-Leibler divergence is the most appropriate statistical measure and should be used for automatic image segmentation of impact damage in CFRP composites.

Lightning strike events pose significant challenges to the structural integrity and performance of composite materials, particularly in aerospace, wind turbine blade, and infrastructure applications. Through a meticulous examination of the state-of-the-art methodologies of laboratory testing and damage predictive modeling, this review elucidates the role of simulated lightning strike tests in providing inputs required for damage modeling and experimental data for model validations. In addition, this review provides a holistic understanding of what is there, what are current issues, and what is still missing in both lightning strike testing and modeling to enable a robust and high-fidelity predictive capability, and challenges and future recommendations are also presented. The insights gleaned from this review are poised to catalyze advancements in the safety, reliability, and durability of composite materials under lightning strike conditions, as well as to facilitate the development of innovative lightning damage mitigation strategies.

In this work, a micromechanics-based model for determining high-temperature thermophysical properties of carbon fiber reinforced polymer (CFRP) composites undergoing pyrolysis is presented. The model accounts for the mass loss and material phase changes. Material phase changes include emergence of the secondary char and gas phases in the polymer matrix. First-order Arrhenius kinetics is used to model polymer pyrolysis. The model also accounts for the temperatureand heating rate-dependent volume fractions of the polymer, char, and pyrolysis gas phases. Temperature- and heating rate-dependent representative volume elements (RVEs) of evolving microstructures were generated and numerical homogenization was performed to determine overall thermophysical properties of the CFRP composites. The micromechanics-based models were embedded into the finite element analysis (FEA)-based multiphysics modeling of lightning strike damage in CFRP composites. Computational studies were performed to analyze lightning strike induced thermal damage in a CFRP composite.

In this work, novel unsupervised machine learning (ML) algorithms for automatic image segmentation for the analysis of the micro-CT data for impact damage assessment in the composite materials have been developed. The algorithms are based on the statistical distances including the Kullback-Leibler divergence, the Helling distance, and the Renyi divergence. The developed algorithms have been applied to the analysis of low velocity impact damage in carbon fiber reinforced polymer (CFRP) composites. The grayscale images from the CT scans of the impacted CFRP specimens have been analyzed to identify and isolate impact damage and optimal statistics-based damage thresholds have been found. The results show that the developed algorithms enable not only an automatic image segmentation, but also deliver statistics-based rigorous damage thresholds.

Adhesive bonding to join fiber reinforced polymer matrix composites holds great promise to replace conventional mechanical attachment techniques for joining composite components. Understanding the behavior of these adhesive joints when subjected to various environmental loads, such as lightning strike, represents an important concern in the safe design of adhesively bonded composite aircraft and spacecraft structures. In the current work, simulated lightning strike tests are performed at four elevated discharge impulse current levels (71.4, 100.2, 141, and 217.8 kA) to evaluate the effects of lightning strike on the mechanical behavior of single lap joints. After documentation of the visually observed lightning strike induced damage, single lap shear tests are conducted to determine the residual bond strength. Post-test visual observation and cross-sectional microscopy are conducted to document the failure modes of the adhesive region. Although the current work was performed on a limited number of specimens, it identified important trends and directions for future more comprehensive studies on lightning strike effects in adhesively bonded composites. It is found that the lightning strike induced damage (extent of the surface vaporization area and the delamination depth) increases as the lightning current increases. The stiffness of the adhesive joints and shear bond strength did not show a clear correlation with the lightning current levels, which could be due to many competing factors, including the temperature rise caused by the lightning strike and the surface conditions of the adherends prior to bonding. The failure modes of the adhesive regions for all specimens demonstrate a mixed mode of adhesive and cohesive failure, which may be due to inconsistent surface characteristics of the adherends before bonding. The energy absorbed during the lap shear tests generally increases as the lightning current increases.

In this work, the role of image segmentation in the analysis of the micro-CT data for the low velocity damage assessment in carbon fiber reinforced polymer (CFRP) composites is discussed. A novel automatic image segmentation method based on the unsupervised learning approach and the Kullback–Leibler divergence is presented. The method has been successfully applied to identify and isolate impact damage in the CFRP composites subjected to the low velocity impact. The results show that the method enables not only an automatic image segmentation, but also delivers a statistics based rigorous damage threshold.

In this paper, an optimization-based computational framework for the spatial tailoring of a metal-ceramic composite panel subjected to high-speed flow is discussed. The framework includes the modeling, evaluation, and optimization of the spatial material grading and thermostructural response of the metal-ceramic composites over a wide range of temperatures. The framework relies on micromechanics and a finite-element analysis (FEA) of representative volume elements (RVEs) to obtain the overall elastic, thermoelastic, and thermal properties of the graded microstructure as functions of temperature and spatial position. The effective thermostructural response of the airframe is analyzed using the FEA. The time-dependent thermal and structural loads are representative of a characteristic high-speed trajectory. Optimal multivariable material distribution is determined numerically using a constrained sequential quadratic programming (SQP) method of surrogate models to evaluate the response at multiple design locations efficiently. Three example cases are presented to showcase the developed framework. In all three example cases, optimal material variation and panel thickness are found such that they reduce the section mass when compared to a benchmark titanium (Ti-6Al-4V) structural skin and Acusill II thermal protection system (TPS) solution. Furthermore, these studies demonstrate that the use of metal-ceramic spatially tailored materials makes excellent material choices for operation in the high-speed environment.

In this paper we discuss the effect of volumetric ablation on the overall elastic properties of the carbon fiber reinforced polymer matrix composite. An Arrhenius type equation describing polymer decomposition was used to determine volume fractions of evolving polymer matrix phases (i.e. polymer, growing pores filled with pyrolysis gases, and char). The effect of the pressure exerted by pyrolysis gases trapped inside the pores was analyzed. Microstructures consisting of carbon fibers (circular inclusions) in the matrix and pores (elliptic inclusions) in the polymer were generated. Temperature dependency was addressed by generating microstructures with different volume fraction of pores, which were calculated from the mass loss model. Two-step numerical homogenization of representative volume elements (RVEs) was performed using finite element analysis (FEA). The developed procedures were applied to calculate temperature dependent (up to 700 K) effective elastic properties of the AS4/3501-6 composite. The results are compared to the existing experimental data and show good agreement.

In this work, a micromechanics-based framework for determining high temperature thermophysical and mechanical properties of CFRPs undergoing thermal decomposition has been further developed. The goal is to develop a framework that enables to investigate the effects of the mass loss and material phase changes (i.e. formation of the secondary char and gas phases), heating rate, decomposition reaction, and pyrolysis pore pressure on the overall material properties. The present work is mainly focused on the investigation of the effects of the heating rate and decomposition reaction. Temperature- and heating rate-dependent volume fractions of the decomposing matrix constituent phases (polymer, char, and pores) are determined from the firstorder Arrhenius kinetics. Representative volume elements (RVEs) with microstructures corresponding to specific temperature and heating rates are created and two-step numerical homogenization of RVEs has been performed to determine overall material properties of the AS4/3501-6 composite in a temperature range up to 900 K. It was determined that overall material properties exhibit strong dependence not only on the temperature but also on the heating rate. An increase in the heating rate shifts initiation of thermal decomposition to the higher temperature. As a result, composite properties are retained at the higher temperatures.

In this paper we discuss the role of different image segmentation methods that are used for the analysis of the micro computed tomography (micro-CT) data of damage in the carbon fiber reinforced polymer (CFRP) composites due to low velocity impact. Segmentation is one of the most critical steps in the image processing of the three dimensional (3D) CT data and accurate assessment of the damage from CT data depends to a great extent on the image segmentation. We have extensively studied low velocity impact damage in the CFRP composites using 3D CT. CFRP textile composite laminates were impacted using an Instron 8200 Dynatup drop-weight impact machine. ZEISS METROTOM 1500 CT scanner was used to evaluate internal impact damage. VGStudio MAX was used for reconstruction of CT images. Different segmentation procedures were used during image processing of the CT images. Differences in the estimates of the damage zone obtained using different segmentation techniques have been assessed.