Search

A novel Regularising Ensemble Kalman filter Algorithm based on the Bayesian paradigm was applied to RTM processes to estimate local porosity and permeability of fibrous reinforcements using measured values of local resin pressure and flow front positions during resin injection. The algorithm allows to detect locations of defects in the preform. It was tested in virtual experiments with two geometries, a two-dimensional rectangular preform and a more complex 3D shape, as well as in laboratory experiments. In both the virtual and laboratory experiments, it was demonstrated that the proposed methodology is able to successfully discover defects and estimate local porosity and permeability with good accuracy. The algorithm also provides confidence intervals for the predictions and estimations of defect probabilities, which are valuable for analysis of the process.

International audience; This paper reports the results of an international benchmark exercise on the measurement of fibre bed compaction behaviour. The aim was to identify aspects of the test method critical to obtain reliable results and to arrive at a recommended test procedure for fibre bed compaction measurements. A glass fibre 2/2 twill weave and a biaxial (±45°) glass fibre non-crimp fabric (NCF) were tested in dry and wet conditions. All participants used the same testing procedure but were allowed to use the testing frame, the fixture and sample geometry of their choice. The results showed a large scatter in the maximum compaction stress between participants at the given target thickness, with coefficients of variation ranging from 38 % to 58 %. Statistical analysis of data indicated that wetting of the specimen significantly affected the scatter in results for the woven fabric, but not for the NCF. This is related to the fibre mobility in the architectures in both fabrics. As isolating the effect of other test parameters on the results was not possible, no statistically significant effect of other test parameters could be proven. The high sensitivity of the recorded compaction pressure near the minimum specimen thickness to changes in specimen thickness suggests that small uncertainties in thickness can result in large variations in the maximum value of the compaction stress. Hence, it is suspected that the thickness measurement technique used may have an effect on the scatter.

© 2020 The Authors 3D woven composites are frequently employed due to their improved through-thickness properties and high damage tolerance compared with laminated composites. Due to the large design space for 3D weave patterns, an in-depth understanding of the relationship between the weave parameters and mechanical properties is essential for the design of these materials. This numerical study investigates the effect of fibre architecture on the mechanical performance of 3D woven composite T-joints under tensile pull-off loading. Six weave pattern variations, subjected to the same preform manufacturing constraint, are designed and numerically analysed, along with another two that have been manufactured and tested for validation previously. Results show a significant architecture dependence in the mechanical responses. Following the design of experiments on weave patterns, the complex architecture-dependant effect is decoupled by two independent variables, yarn path entanglement and yarn path crossover. The study also provides design recommendations for 3D woven T-joint reinforcements under tensile pull-off loading.

This paper presents a comprehensive review of the scientific and engineering contributions of Prof. Stepan V. Lomov in a career spanning over 3 decades. Starting from his early work on the modelling of dry textiles, different aspects of composite materials have benefited from his inspiring contribution. His wide scientific curiosity drove the research on several topics, covering permeability, forming, damage, fatigue, nano-engineered composites, as well as development of analytical and numerical models at different scales. The paper presents a full spectrum of his scientific career aiming to link his achievements and the future research on composite materials.

Axial flow through gaps between aligned straight yarns with realistic cross-sectional shapes, described by power-ellipses, was analysed numerically. At a given fibre volume fraction, equivalent gap permeabilities have a maximum at minimum size of elongated tapering parts of the gap cross-section and a ratio of gap width to height near 1. When the yarn spacing is given in addition to the fibre volume fraction, calculated maximum and minimum values for the equivalent permeability of inter-yarn gaps, which occur at near-rectangular and lenticular cross-sections, differ by factors of up to 3.3. Novel approximations for the shape factor and the hydraulic diameter in Poiseuille flow were derived as a function of the fibre volume fraction, the yarn cross-sectional aspect ratio and the exponent describing the shape of the power-elliptical yarn cross-section. This allows the equivalent gap permeability to be predicted with good accuracy for any fibre volume fraction and yarn cross-section.

Resin transfer moulding is one of several processes available for manufacturing fibre-reinforced composites from dry fibre reinforcement. Recently, dry reinforcements made with Automated Dry Fibre Placement have been introduced into the aerospace industry. Typically, the permeability of the reinforcement is assumed to be constant throughout the dry preform geometry, whereas in reality, it possesses inevitable uncertainty due to variability in geometry. This uncertainty propagates to the uncertainty of the mould filling and the fill time, one of the important variables in resin injection. It makes characterisation of the permeability and its variability an important task for design of the resin transfer moulding process. In this study, variability of the geometry of a reinforcement manufactured using Automated Dry Fibre Placement is studied. Permeability of the manufactured preforms is measured experimentally and compared to stochastic simulations based on an analytical model and a stochastic geometry model. The simulations showed that difference between the actual geometry and the designed geometry can result in 50% reduction of the permeability. The stochastic geometry model predicts results within 20% of the experimental values.

A common method to fabricate net-shaped three-dimensional (3D) woven preforms for composite T-joints is to weave flat 3D preforms via a standard weaving machine with variation in binder yarn path and then separate the preform in the form of a bifurcation. Folding introduces fiber architecture deformation at the 3D woven bifurcation area. In this paper, a geometric modeling approach is proposed to represent the realistic fiber architecture, as a preprocessor for finite element analyses to predict composite structural performance. Supported by X-ray micro-computed tomography (µCT), three important deformation mechanisms are observed including yarn stack shifting, cross-section bending, and cross-section flattening resulting from the folding process. Furthermore, a set of mathematical formulae for simulation of the deformations in the junction region are developed and satisfactory agreement is observed when compared with μCT scan results.

Although good progress was made by two international benchmark exercises on in-plane permeability, existing methods have not yet been standardized. This paper presents the results of a third benchmark exercise using in-plane permeability measurement, based on systems applying the radial unsaturated injection method. 19 participants using 20 systems characterized a non-crimp and a woven fabric at three different fiber volume contents, using a commercially available silicone oil as impregnating fluid. They followed a detailed characterization procedure and also completed a questionnaire on their set-up and analysis methods. Excluding outliers (2 of 20), the average coefficient of variation (c v ) between the participant´s results was 32% and 44% (non-crimp and woven fabric), while the average c v for individual participants was 8% and 12%, respectively. This indicates statistically significant variations between the measurement systems. Cavity deformation was identified as a major influence, besides fluid pressure/viscosity measurement, textile variations, and data analysis. Fil: May, D.. Institut Für Verbundwerkstoffe Gmbh; Alemania Fil: Aktas, A.. National Physical Laboratory; Reino Unido Fil: Advani, S.G.. University Of Delaware; Estados Unidos Fil: Berg, D. C.. Technische Universität Clausthal; Alemania Fil: Endruweit, A.. University of Nottingham; Estados Unidos. Science and Technology Facilities Council of Nottingham. Rutherford Appleton Laboratory; Reino Unido Fil: Fauster, E.. Montanuniversitat Leoben; Austria Fil: Lomov, S. V.. Katholikie Universiteit Leuven; Bélgica Fil: Long, A.. University of Nottingham; Estados Unidos. Science and Technology Facilities Council of Nottingham. Rutherford Appleton Laboratory; Reino Unido Fil: Mitschang, P.. Institut Für Verbundwerkstoffe Gmbh; Alemania Fil: Abaimov, S.. Skolkovo Institute Of Science And Technology; Rusia Fil: Abliz, D.. Technische Universität Clausthal; Alemania Fil: Akhatov, I.. Skolkovo Institute Of Science And Technology; Rusia Fil: Ali, M. A.. Khalifa University Of Science And Technology; Emiratos Arabes Unidos Fil: Allen, T. D.. University of Auckland; Nueva Zelanda Fil: Bickerton, S.. University of Auckland; Nueva Zelanda Fil: Bodaghi, M.. Institute Of Science And Innovation In Mechanical And Industrial; Portugal Fil: Caglar, B.. Koç Üniversitesi; Turquía Fil: Caglar, H.. Koç Üniversitesi; Turquía Fil: Chiminelli, A.. Instituto Tecnologico de Aragon; España Fil: Correia, N.. Institute Of Science And Innovation In Mechanical And Industrial; Portugal Fil: Cosson, B.. Imt Lille Douai; Francia Fil: Danzi, M.. Eth Zürich; Suiza Fil: Dittmann, J.. Universität Stuttgart; Alemania Fil: Ermanni, P.. Eth Zürich; Suiza Fil: Francucci, Gaston Martin. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Mar del Plata. Instituto de Investigaciones en Ciencia y Tecnología de Materiales. Universidad Nacional de Mar del Plata. Facultad de Ingeniería. Instituto de Investigaciones en Ciencia y Tecnología de Materiales; Argentina Fil: George, A.. University Brigham Young; Estados Unidos Fil: Grishaev, V.. Skolkovo Institute Of Science And Technology; Rusia Fil: Hancioglu, M.. Koç Üniversitesi; Turquía Fil: Kabachi, M. A.. Eth Zürich; Suiza Fil: Kind, K.. Universitat Technical Zu Munich; Alemania Fil: Deléglise Lagardère, M.. Imt Lille Douai; Francia Fil: Laspalas, M.. Instituto Tecnologico de Aragon; España Fil: Lebedev, O. V.. Skolkovo Institute Of Science And Technology; Rusia Fil: Lizaranzu, M.. Instituto Tecnologico de Aragon; España Fil: Liotier, P. J.. Université de Lyon; Francia Fil: Middendorf, P.. Universität Stuttgart; Alemania Fil: Morán, J.. Universidad Nacional de Mar del Plata; Argentina Fil: Park, C. H.. Imt Lille Douai; Francia Fil: Pipes, R. B.. Purdue University; Estados Unidos Fil: Pucci, M. F.. Université de Montpellier; Austria Fil: Raynal, J.. Groupe Institut de Soudure; Francia Fil: Rodriguez, E. S.. Universidad Nacional de Mar del Plata; Argentina Fil: Schledjewski, R.. Montanuniversitat Leoben; Austria Fil: Schubnel, R.. Groupe Institut de Soudure; Francia Fil: Sharp, N.. Purdue University; Estados Unidos Fil: Sims, G.. National Physical Laboratory; Reino Unido Fil: Sozer, E. M.. Koç Üniversitesi; Turquía Fil: Sousa, P.. Katholikie Universiteit Leuven; Bélgica Fil: Thomas, J.. Khalifa University Of Science And Technology; Emiratos Arabes Unidos Fil: Umer, R.. Khalifa University Of Science And Technology; Emiratos Arabes Unidos Fil: Wijaya, W.. University of Auckland; Nueva Zelanda Fil: Willenbacher, B.. Institut Für Verbundwerkstoffe Gmbh; Alemania Fil: Yong, A.. National Physical Laboratory; Reino Unido Fil: Zaremba, S.. Universitat Technical Zu Munich; Alemania Fil: Ziegmann, G.. Technische Universität Clausthal; Alemania

This paper reports results on the shielding effectiveness parameter of laminated epoxy composites with carbon fibre reinforcements. Measurements of shielding effectiveness were carried out with a coaxial transmission line testing chamber according to ASTM 4935 standard and epoxy-matrix composites with continuous carbon-fibres were proven to be an excellent electromagnetic interference shielding material, where a composite slab made of 4 layers of prepregs provided more than 99.9% of electromagnetic attenuation. It was found that the reflection mechanism of the shielding material was mainly influenced by the fibre volume ratio, and that an increase in the number of layers of the composite resulted in higher shielding effectiveness due to a greater absorption mechanism. Calculations of the shielding effectiveness parameter of the material used were made by means of commercial electromagnetic simulation tools, having determined experimentally the overall resistivity of the composite. The findings presented in this work suggest that in presence of a greater number of interfaces at different impedance the separate modelling of matrix and fibres at mesoscopic scale must be taken into account.

To understand the influence of the fibre architecture of 3D woven composite T-joints on mechanical performance, as well as the benefits that 3D woven T-joints can offer over the equivalent 2D laminates, experimental testing is performed on two types of 3D woven T-joint with only weave variation at the junction, and one type of 2D woven laminate T-joint. A quasi-static tensile pull-off loading is selected in this work as this out-of-plane load case is one of the typical loading conditions for such T-joint structures. The significant advantages of 3D woven composite T-joints in terms of ultimate strength and damage tolerance over the 2D alternative were identified in the testing. More importantly, this work showed that variation in the fibre architecture can considerably enhance properties such as delamination resistance and total energy absorption to failure, as well as increasing slightly the stiffness and initial failure load. This experimental assessment has demonstrated that using 3D woven reinforcements is an effective way to improve the load-bearing capability of composite T-joints over laminates, and also that this improvement could be optimised with regard to fibre architecture.

Meso-scale unit cell models are often used to simulate mechanical behaviour of textile composites. Apart from reliable ways to create meso-scale geometries, such simulations require reliable meshing algorithms. While the former is made possible via dedicated textile pre-processors or high-fidelity weaving simulations, the meshing remains quite problematic for complex textiles and geometries. Even though, with a lot of user input, it is possible to create very complex meshes using meshing pre-processors, this approach remains infeasible for cases when a large number of models need to be analysed. This paper presents a meshing approach based on the combination of local octree-refinement with surface smoothing. This allows nearly conformal meshes to be generated for geometries of any complexity which achieve accuracy comparable to that of conformal meshes. A range of unit cells was analysed using the new approach and it was shown that the error in local stresses is within 10% of the reference solution and the average error is below 7%. It was found that the computational cost of the analysis using the new meshing technique is not considerably higher than for an analysis which uses a conventional conformal mesh yet the new approach allows analysis of any geometry.

Many technical applications of woven fabric are subject to increasing high pressure from air transport through the fabric. The through-thickness permeability (TP) of woven materials exhibits a dynamic response to increased air pressure. This paper presents an analytical model for predicting the steady TP of woven fabric. The approach was based on Darcy’s law and the Poiseuille equation, using the flow boundary of an idealized plain-weave unit cell. The unit cell model consists of a gradual converging-diverging (GCD) duct with a rectangular cross-section. Further, the dynamic TP of the GCD duct was established analytically as a function of increasing pressure, which correlates to the separation of air flow from the GCD duct wall. Air flow separation from the duct wall led to a quadratic relationship between the increasing pressure and air flow velocities. This dynamic TP and air flow nonlinearity were simulated numerically in the computational fluid dynamics solver CFX. Five GCD ducts under increasing air pressure were analyzed numerically and analytically. The comparison showed good agreement between the proposed analytical model and the CFD simulation, with a maximum error up to 12%. A sensitivity study showed that an increase in porosity or a decrease in the thickness of weave materials could result in a larger dynamic TP value.

3D composite materials are characterized by complex internal yarn architectures, leading to complex deformation and failure development mechanisms. Net-shaped preforms, which are originally periodic in nature, lose their periodicity when the fabric is draped, deformed on a tool, and consolidated to create geometrically complex composite components. As a result, the internal yarn architecture, which dominates the mechanical behaviour, becomes dependent on the structural geometry. Hence, predicting the mechanical behaviour of 3D composites requires an accurate representation of the yarn architecture within structural scale models. When applied to 3D composites, conventional finite element modelling techniques are limited to either homogenised properties at the structural scale, or the unit cell scale for a more detailed material property definition. Consequently, these models fail to capture the complex phenomena occurring across multiple length scales and their effects on a 3D composite’s mechanical response. Here a multi-scale modelling approach based on a 3D spatial Voronoi tessellation is proposed. The model creates an intermediate length scale suitable for homogenisation to deal with the non-periodic nature of the final material. Information is passed between the different length scales to allow for the effect of the structural geometry to be taken into account on the smaller scales. The stiffness and surface strain predictions from the proposed model have been found to be in good agreement with experimental results. The proposed modelling framework has been used to gain important insight into the behaviour of this category of materials. It has been observed that the strain and stress distributions are strongly dependent on the internal yarn architecture and consequently on the final component geometry. Even for simple coupon tests, the internal architecture and geometric effects dominate the mechanical response. Consequently, the behaviour of 3D woven composites should be considered to be a structure specific response rather than generic homogenised material properties.

A stochastic cure simulation approach is developed and implemented to investigate the influence of fibre misalignment on cure. Image analysis is used to characterize fiber misalignment in a carbon non-crimp fabric. It is found that variability in tow orientation is significant with a standard deviation of 1.2°. The autocorrelation structure is modeled using the Ornstein-Uhlenbeck sheet and the stochastic problem is addressed by coupling a finite element model of cure with a Monte Carlo scheme. Simulation of the cure of an angle shaped carbon fiber-epoxy component shows that fiber misalignment can cause considerable variability in the process outcome with a coefficient of variation in maximum residual stress up to approximately 2% (standard deviation of 1 MPa) and qualitative and quantitative variations in final distortion of the cured part with the standard deviation in twist and corner angle reaching values of 0.4° and 0.05° respectively. POLYM. COMPOS., 2015. © 2015 The Authors Polymer Composites published by Wiley Periodicals, Inc. on behalf of Society of Plastics Engineers

Through-thickness permeability (TTP) is one primary property of technical textiles used in air-related applications, such as filtration and protection. The TTP depends on the textile geometrical factors and usually varies according to the test conditions. In this article, the effect of low air pressure compression (LPC) on TTP of woven fabric was investigated. Nine woven fabrics were measured for the relationships of LPC and thickness, LPC and fabric in-plane dimensions, air pressure drop (APD) and air velocity, as well as LPC and fabric TTP. A dramatic decrease of woven fabric thickness was found below the APD value of 200 Pa and less decreased thickness was observed with a continued increase of APD. The variation of fabric in-planar dimensions was found to be negligible during LPC. The plot relationship of the APD and measured air velocity was presented in linearity for most fabric samples. The fabric TTP showed a linear proportion to the fabric thickness, indicating the fabric to be more permeable with the increase of thickness. A sensitivity study showed an evident difference between using fabric constant and decreased (LPC) thickness in calculating TTP, disclosing the importance of compression in fabric TTP evaluation.

When a woven fabric is subject to a normal uniform loading, its properties such as tightness and through-thickness permeability are both altered, which relates to the fabric out-of-plane deformation (OPD) and dynamic permeability (DP). In this article, fabric OPD is analytically modelled through an energy minimisation method, and corresponding fabric DP is established as the function of loading and fabric-deformed structure. The total model shows the permeability a decrease for tight fabric and an increase for loose fabric when the uniform loading increases. This is verified experimentally by fabric OPD, static and dynamic permeabilities. Experimental tests for both permeabilities showed good agreement with the corresponding predictions, indicating the fact that tight fabric becomes denser and loose fabric gets more porous during OPD. A sensitivity study showed that an increase of fabric Young’s modulus or a decrease of fabric test radius both lead to an increase of DP for tight fabric and opposite for loose fabric. The critical fabric porosity and thickness were found for inflexion of fabric DP trend during the OPD, which contributes to the optimum design of interlacing structure applied to protective textiles and composites.

Composites manufacturing involves many sources of uncertainty associated with material properties variation and boundary conditions variability. In this study, experimental and numerical results concerning the statistical characterisation and the influence of inputs variability on the main steps of composites manufacturing including process-induced defects are presented and analysed. Each of the steps of composite manufacturing introduces variability to the subsequent processes, creating strong interdependencies between the process parameters and properties of the final part. The development and implementation of stochastic simulation tools is imperative to quantify process output variabilities and develop optimal process designs in composites manufacturing.

For a 3D orthogonal carbon fibre weave, geometrical parameters characterising the unit cell were quantified using micro-Computed Tomography and image analysis. Novel procedures for generation of unit cell models, reflecting systematic local variations in yarn paths and yarn cross-sections, and discretisation into voxels for numerical analysis were implemented in TexGen. Resin flow during reinforcement impregnation was simulated using Computational Fluid Dynamics to predict the in-plane permeability. With increasing degree of local refinement of the geometrical models, agreement of the predicted permeabilities with experimental data improved significantly. A significant effect of the binder configuration at the fabric surfaces on the permeability was observed. In-plane tensile properties of composites predicted using mechanical finite element analysis showed good quantitative agreement with experimental results. Accurate modelling of the fabric surface layers predicted a reduction of the composite strength, particularly in the direction of yarns with crimp caused by compression at binder cross-over points.

Jute yarns were treated by tap water with and without tension at room temperature for 20 minutes and then dried. Fibre and yarn strength were measured before and after treatment. Unidirectional (UD) composites were made by both treated and untreated yarns with and without applying hydroxyethyl cellulose (HEC) as size material. Water-treated jute yarns without tension and composites made of those yarns showed decreased strength, and water treated jute yarns with tension and composites made of those yarns showed increased strength with respect to raw yarns and composites made of raw yarns. However, no specific trend was noticed for fibre tensile strength and tensile modulus. HEC sized yarns showed up to 12% higher failure load with respect to unsized yarns, and composites made of HEC sized yarns showed up to 17% and 12% increase in tensile strength and tensile modulus, respectively, compared to composites made of similar types of unsized yarns.

A solution for the transverse thermal conductivity (K e ) of unidirectional fibre arrays, quadratic and hexagonal, is developed analytically. The solution integrates the thermal conductivity of fibre (K f ) and fluid (K m ) based on electricity analogy without thermal contact resistance (R c ) at the fibre/fluid interface. The expression K e is a function of K f and K m , as well as of the fibre volume fraction (V f ). In this article, K e values of four composites were predicted and verified by computational fluid dynamics (CFD) simulations. The results showed good agreement when the ratio of K f /K m is close to 1. An increase in the ratio or V f gives poorer agreement owing to the local temperature gradient at the fibre/fluid interface. CFD simulation also showed that K e is decreasing as the R c value increases.

For fibre bundles with non-uniformly distributed filaments, probabilities for the occurrence of defined local filament arrangements, determining geometrical constants for permeability calculation, were described as functions of the local fibre volume fraction. Because of changing inter-filament gap geometries, the local transverse permeability decreases discontinuously with increasing local fibre volume fraction, while the local axial permeability decreases continuously. Axial bundle permeabilities, derived from log-normally distributed local permeabilities, tend to be higher than predictions for uniform filament arrangements, while transverse bundle permeabilities tend to be lower. This implies that assuming uniformity of filament arrangements is inappropriate for realistic modelling of bundle properties. For fibre bundles with rectangular cross-section, the transverse permeability decreases in width-direction and increases in thickness-direction with increasing aspect ratio of the bundle cross-section, while the axial permeability is not affected. This is of high practical relevance, in particular for spread-tow reinforcements with high aspect ratios.

Highly structured and hierarchal textiles may be modelled at three scales, ranging from micro modelling of individual fibres within a yarn via meso-scale modelling of a fabric unit cell through to macro-scale modelling of a whole product. The performance of the product can be designed and optimised effectively by variation of the structure at every level. This paper presents an integration of the fibre–yarn, yarn–fabric models. The micro-scale model is developed using an energy-based argument combined with elastica methods to incorporate non-linear frictional behaviour. The meso-scale model is developed using the finite element method, with the yarn modelled as a continuum. The output of the micro-scale model is used to determine constitutive materials properties at the meso-scale model for fabric deformation. The main features of this modelling approach are that many critical design parameters of a yarn and a fabric are being addressed and the friction is incorporated at each scale since inter-fibre and inter-yarn friction may be directly affected by finishing treatments. This paper shows how predictions at the micro scales can pass up the hierarchy to the meso scale to enable designers/engineers to understand and quantify how changes in yarn properties and fabric architectures influence the performance of the fabric, thereby facilitating the design and optimisation of fabric properties by variation of its constituent structural components.

Technologies for automated fibre lay-up have proven their usefulness in composites manufacturing. Further development of the technologies, such as Automated Dry Fibre Placement (ADFP), allow further reduction of waste and increase of the design space through tow steering which enables creation of composites with tailored properties. Tow steering is, however, limited by possible defects such as wrinkles which result from mismatch of fibre length and steering path. This paper addresses wrinkle formation at different steering radii and provides a closed-form solution for the problem. Experimental results are used for estimation of the model parameters and validation of the model. An analytical framework is used to explore effects of processing parameters on defect formation. It was found that the tack stiffness has the greatest influence on defect formation. Parametric studies showed that increase of the temperature within the admissible temperature window can improve the tack properties and hence improve the lay-up.

The paper proposes a novel type of stiffener designed to bear bending loads by exploiting internal pressure effects. The stiffener is made of two adjacent thin-walled pipes (r/t≥50) jointed with a connecting strip. Such a structure is shown to have higher performance against buckling failure compared to a single pipe and its geometry allows for good exploitation of internal pressurisation. The study is conducted by using the FEA software ANSYS and the analysis technique is the linear perturbation buckling analysis. Internal pressure ranges from 0 to 1.4 MPa. The buckling mechanisms are observed for a set of models with different values of length, wall thickness and geometric variation of the cross-section. It is shown that two different buckling modes can take place. However, for a given geometry, the level of pressure can alter the behaviour and lead to one mode rather than the other one. Potential of the presented structure is maximised by the use of high performance materials and a possible aerospace engineering application is presented.

This work presents a detailed finite element approach for textile fabrics for mechanical analysis and simulation by combining yarn properties, inter-yarn interaction and fabric structures. The fabric geometries are generated using the TexGen geometric modelling schema developed at the University of Nottingham. This geometric model of the fabric is interfaced with a mechanics modelling environment, ABAQUS, to predict its mechanical properties. The major advantages over other approaches are that all issues arising from textile mechanics modelling are considered, physically measured data are used as inputs and the traditional requirement of user intervention during model creation is removed. The developed model and its implementation are validated using the Kawabata Evaluation System for Fabrics on commercial fabrics subjected to variety of external loadings. The presented validation shows good agreement between the simulations utilising the finite element model and the experiments. The modelling approach is...

An automated approach (TexGen) for modeling the geometry of textile structures is presented. This model provides a generic approach to the description of yarn geometry and yarn interlacement for all types of weaving. One feature of this model is that the shape and size of the cross sections may change locally; this is exploited in the functions for interference correction, which modify the textile according to geometric considerations to avoid penetration of yarns. Another feature of this model is that it acts as a pre-processor for finite element simulations by generating a mesh, definition of contact, materials orientation and boundary conditions, thus providing an automatic procedure. This paper describes the modeling techniques, algorithms and concepts implemented in TexGen and examines the functionality of their implementation for a range of two-dimensional/three-dimensional commercial fabrics. Comparisons between the images of real fabrics and modeled fabric structures confirm that the software is capable of modeling sophisticated fabric architectures, including twisted yarns with varied yarn cross sections. Accurate input measurements of fabric geometry are critical for successful results. The paper also discusses directions for further development of the approach to overcome current limitations.

Variability of tow orientation is unavoidable for biaxial engineering fabrics and their composites. Since the mechanical behaviour of these materials is strongly dependent on the fibre direction, variability should be considered and modelled as exactly as possible for more realistic estimation of their forming and infusion behaviour and their final composite mechanical properties. In this study, a numerical code, ‘VariFab’, has been written to model realistic full-field variability of the tow directions across flat sheets of biaxial engineering fabrics and woven textile composites. The algorithm is based on pin-jointed net kinematics and can produce a mesh of arbitrary perimeter shape, suitable for subsequent computational analysis such as finite element forming simulations. While the shear angle in each element is varied, the side-length of all unit cells within the mesh is constant. This simplification ensures that spurious tensile stresses are not generated during deformation of the mesh during forming simulations. Variability is controlled using six parameters that can take on arbitrary values within certain ranges, allowing flexibility in mesh generation. The distribution of tow angles within a pre-consolidated glass–polypropylene composite and self-reinforced polypropylene and glass fabrics has been characterised over various length scales. Reproduction of the same statistical variability of tow orientation as in these experiments is successfully achieved by combining the VariFab code with a simple genetic algorithm.

Woven fabric permeability is relevant to many applications, such as airbags, textile composites processing, paper making and air and water filtration. This paper proposes an analytical model to predict through-thickness fabric permeability based on viscous and incompressible Hagen–Poiseuille flow. The flow is modeled through a unit cell of fabric with a smooth fluid channel at the center with slowly varying cross-section. The channel geometry is determined by yarn spacing, yarn cross-section and fabric thickness. The shape of the channel is approximated by a parabolic function. Volumetric flow rate ( Q) is formulated as a function of pressure drop and flow channel geometry for woven fabric. The permeability ( K) is calculated thereafter according to Darcy’s law. The air permeability of nine different fabrics has been measured to verify the analytical model. A sensitivity study was carried out to understand how geometric parameters influence the fabric permeability. The analytical model shows very close agreement with the experimental data: within 5% for most fabrics. The sensitivity study on permeability indicates the importance of flow channel geometry in obtaining accurate predictions.

Dynamic permeability is relevant to textile applications subjected to fluid/gas flow under high pressure, such as automotive airbags, wearable airbags and parachute fabrics. Dynamic permeability can be determined when a porous medium is tested under transient pressure conditions. This paper utilizes a reliable approach to measure and characterize dynamic permeability for woven fabrics. The experimental principle is based on the ideal gas law and the non-linear Forchheimer equation. Compared with static permeability measured under a constant low pressure, the dynamic permeability is an intrinsic property determined by change of fabric geometry and structure due to a high-pressure load. The pressure-induced deformation is identified, including effects on fiber and yarn arrangement, yarn porosity and fabric thickness. The level of deformation is a function of the number of fabric layers and initial pressure drop. The experimental results show that the dynamic permeability is higher than the static permeability for loose fabric, while it is lower for tight fabrics. For tight fabric, more fabric layers and a lower initial pressure can reduce the difference between the static and the dynamic permeability. Analytical models are used to explain and predict both static and dynamic permeability.

In this article, a numerical method to predict textile geometry is derived using a technique based on finite elements (FEs). A geometric modeling package is used to represent an initial geometry of the yarns within the textile. The yarn mid-surface is then represented using plate elements, with the yarn thickness and cross-section being reconstructed from this mid-surface. The bending and tensile aspects of the yarn behavior are represented by separate features of the plate elements and the total energy for the system is minimized. Contacts are modeled using a penalty method, where the contact force is proportional to penetration distance. Once geometry correction has been achieved by solving the FE problem, the geometric model of the textile is corrected to take into account the predicted movements of the yarns. For validation purposes, the method is applied to two-dimensional (2D) and three-dimensional (3D) weaves and compared against images of the real fabrics. Agreement between predictions and images is good for the 3D weave and excellent for the 2D weave.

A predictive approach to modelling the forming of viscous textile composites has been implemented in two finite element codes; Abaqus Standard™ and Abaqus Explicit™. A multi-scale energy model is used to predict the shear force–shear angle–shear rate behaviour of viscous textile composites, at specified temperatures, using parameters supplied readily by material manufacturers, such as fibre volume fraction, weave architecture and matrix rheology. The predictions of the energy model are fed into finite element simulations to provide the in-plane shear properties of two different macro-scale constitutive models implemented in the finite element codes. The manner of coupling predictions of the multi-scale energy model with the macro-scale models is shown to affect the rate-dependent material response in the simulations. These coupling methods are evaluated using picture frame test simulations.

In this study, the local fluid concentration during in-plane impregnation of various reinforcement fabrics was mapped employing Magnetic Resonance Imaging (MRI). Repeated MRI scans during intermittent injection allowed the development of meso-scale dry spots to be tracked individually. The dry spots were found to be stationary and to decrease in size during impregnation. Changes in local fluid concentration within fibre bundles, reflecting the micro-scale void content, indicate that, behind the flow front, a zone of partial saturation travels through the specimen. For each type of specimen, this has a characteristic width. 3D maps of the fluid content in the specimens provided qualitative verification of models for the impregnation behaviour of specific fabrics. The results show the potential of MRI to provide unique 3D information on the local fluid content and the formation and development of dry spots.

This paper provides an overview of TexGen, the open source software package for 3D modelling of textiles and their composites developed at the University of Nottingham. The underlying modelling theory is briefly discussed followed by descriptions of applications utilising TexGen in the fields of textile mechanics, textile composite mechanics and permeability. The limitations and further development of the approach are also considered.

In this international permeability benchmark exercise, in-plane permeability data for two reinforcement fabrics, obtained using a total of 16 different experimental procedures, were compared. Although, for each procedure, the results appear consistent, different procedures result in a scatter of up to one order of magnitude in principal permeability values for each fabric at any given fibre volume fraction. The ratio of the principal permeability values varies by factors of up to 2. While experimental uncertainties and variability of the specimens affect the scatter in results for any single series of experiments, it is suspected that the main source of scatter in results from different procedures is related to human factors. Aiming at standardisation of measurement methods and interchangeability of results, "good practice" guidelines will be formulated in order to eliminate sources of scatter. (C) 2011 Elsevier Ltd. All rights reserved.

Magnetic Resonance Imaging techniques were found to be suitable for mapping the fluid distribution in impregnated textile fabrics. They allowed the state of impregnation of E-glass and carbon fibre fabrics of different architectures with a test fluid (engine oil) to be identified and dry spots to be detected. An Ultra-Short Echo-Time technique was found to be relatively robust to signal loss caused by relaxation processes, which are related to the dispersion of the fluid in the fabric and local discontinuities in magnetic susceptibility. On a 3 T scanner, 3D images of specimens with dimensions of 140 mm × 90 mm × 4.7 mm at an isotropic resolution of 0.5 mm were acquired at scan times of 21 min. However, radio-frequency eddy currents may be induced if the fabrics are conductive and result in partial cancellation of the signal, rendering it undetectable inside the sample.

For multilayer preforms from three different triaxial carbon fibre braids with bias angles 45°, 60° and 70° at fibre volume fractions of 0.54 and 0.59, the in-plane permeability was characterised experimentally. A finding of high practical relevance is that, at a bias angle of approximately 55°, the principal flow direction switches from the preform 0° direction to the preform 90° direction. For different zones of characteristic yarn arrangement in the braid unit cell, the local permeability was modelled as a function of bias angle, global fibre volume fraction, and geometrical yarn parameters. The global braid permeability was derived from numerical flow simulation based on simplified 2D models of the textile architecture and local permeabilities, taking into account effects of nesting between adjacent layers. Calculated trends for the permeability as a function of the braid angle reproduce the experimental results qualitatively well. Quantitatively, agreement depends on the value of a geometry parameter, which decreases with increasing fibre volume fraction.

TexGen is open source software developed at the University of Nottingham for the geometric 3D modelling of textiles and textile composites. It has a large number of users worldwide and underpins a significant number of research publications. While many users make simplifying assumptions about the structure of a textile, in reality the internal geometry of a textile or textile composite is complex. Capturing this complexity is vital for the prediction of properties such as permeability and mechanical failure. Examples will be given of the characterisation of a material and how the complex features are captured and implemented in TexGen, making use of functionality such as the ability to vary the cross-sectional shape along the length of a yarn. The effect on prediction of properties as a model is refined will be demonstrated. Recent additions to the software will also be highlighted. Laminated structures can be quickly and easily constructed from a selection of textiles and several nesting options are available. A new rotate textile option can then be used to create laminates with varying ply angles. Where the unit cell is also rotated, appropriate periodic boundary conditions have been implemented and are automatically generated in an ABAQUS input file. A new feature is described which generates a TexGen model from a weave pattern file. Future developments of this may improve accessibility of the software to the weaving community. The generation of a pattern draft output from the TexGen model is also described.

For three 3D woven carbon fiber reinforcements with different architectures, the compressibility, geometrical structure, and permeability were studied. At low levels of compression, the thickness of an angle-interlock weave is reduced mainly by local reduction of the height of inter-bundle voids and permanent reordering of the fiber bundles. At higher compression levels, bundle compaction is dominant. For an orthogonal weave fabric, the main compression mechanism is compaction of the fiber bundles. The in-plane permeability is modeled by superimposing a disturbance, caused by the binder yarns, to the permeability of a regular layered fiber structure. It is characterized mainly by the dimensions of the inter-bundle voids in each layer, and the pattern and dimensions of the binder yarns. Because of the dependence on the inter-bundle void dimensions, the fabric compression behavior is reflected. The through-thickness permeability is determined by flow-enhancing channels in the structure of the reinforcement, formed around the binder yarns. It is modeled as a function of the number of binder yarns per unit surface area, and the dimensions and orientation of the binder yarns.

The impregnation stage of the Resin Transfer Moulding process can be simulated by solving the Darcy equations on a mould model, with a ‘macro-scale’ finite element method. For every element, a local ‘meso-scale’ permeability must be determined, taking into account the local deformation of the textile reinforcement. This paper demonstrates that the meso-scale permeability can be computed efficiently and accurately by using meso-scale simulation tools. We discuss the speed and accuracy requirements dictated by the macro-scale simulations. We show that these requirements can be achieved for two meso-scale simulators, coupled with a geometrical textile reinforcement modeller. The first solver is based on a finite difference discretisation of the Stokes equations, the second uses an approximate model, based on a 2D simulation of the flow.

A three-dimensional (3D) micromechanical study has been performed in order to investigate local damage in unidirectional (UD) composite materials with epoxy resin under transverse tensile loading. In particular the effect of different mechanical properties of a 3D interphase within the hexagonal array RVE have been considered and effects of thermal residual stress arising during the curing process have been accounted for in this study. To examine the effect of interphase properties and residual stress on failure, a study based on the temperature-dependent properties of matrix and interphase and a stiffness degradation technique has been used for damage analysis of the unit cell subjected to mechanical loading. Results indicate a strong dependence of damage onset and its evolution from the different interphase properties within the RVE (representative volume element). Moreover, predicted mechanical properties, damage initiation and evolution are also clearly influenced by the presence of residual stress. Numerical results and experimental data (in the literature) have also shown an interesting agreement.

It is widely accepted that the key deformation mechanisms during forming of viscous textile composite (prepreg) sheets are in-plane shear and out-of-plane bending. This paper focuses on the bending deformation mechanism, including experimental characterisation and theoretical modelling of bending behaviour during viscous composite forming. Experimental measurements are obtained by means of a large-displacement buckling test at a variety of displacement rates and temperatures. Some important aspects, such as viscoelastic behaviour, are also investigated. A bending model based on elastic theory combined with uniaxial continuum theory for ideal fibre-reinforced fluids for viscous shear deformation has been developed, using material parameters obtained from industrial manufacturers as input data, such as composite geometry, fibre properties, fibre volume fraction and matrix rheology. Model predictions demonstrate that the model can capture the main characteristics of material properties, such as rate dependence. This bending model can be used in formability analysis for viscous unidirectional composite materials, and might be applied in a finite element forming simulation to account for the bending stiffness.

An area of interest in the automated manufacture of composite components is the prediction in real-time of the deformed shape of a textile reinforcement in 3D space during robotic handling operations. The deformed shape can be used to guide robotic end-effectors to ensure accurate fabric placement and avoid collisions. In this paper, a nonlinear mathematical model using large deflection plate and shell theories is presented. The model is able to predict the 3D deformed shape of limp sheet materials being picked-up by multi robotic grippers for three boundary conditions. The main factors affecting the deformation behaviour of the sheet during the operation are identified and analysed, and the contributions of different energies during deformation are presented in detail. Good agreement is obtained when comparing the solutions of the model with FE simulation results. This study demonstrates the possibility of developing a modelling capability for material on-line response in automatic flexible material handling.

A three dimensional (3D) micromechanical study has been performed in order to investigate local damage in UD composite materials under transverse and longitudinal tensile loading. In particular, the influence of non-uniform distribution of fibres in RVEs (representative volume element) with a hexagonal packing array and the effects of thermal residual stresses has been investigated. To examine the effect of inter-fibre spacing and residual stress on failure, a study based on the Maximum Principal Stress failure criterion and a stiffness degradation technique has been used for damage analysis of the unit cell subjected to mechanical loading. Results indicate a strong dependence of damage onset and its evolution from the fibres position within the RVE. Predicted mechanical properties, damage initiation and evolution are also clearly influenced by the presence of residual stress.

New experimental set-ups are presented for measuring the pressure profile and fill-times in the Vacuum Infusion (VI) process. In these set-ups, the injection can either be from one of the mould faces (resulting in a rectilinear flow) or from a central port (resulting in a radial flow). From these measurements, the validity of previously reported analytical formulations is investigated. At the start of injection, the experimental results show a marked difference from analytical predictions. However, with flow progression, they change to match with analytical predictions. This phenomenon has not been observed previously and its analysis enhances the current understanding of the process physics, mainly the impact of compliance on the reinforcement thickness and flow progression.

In his fin-de-siècle Morocco travelogue Mogreb-el-Acksa, Robert Cunninghame Graham refused and traversed Orientalist literary discourse. This essay explains how remarkable this refusal and traversal was, given Cunninghame Graham's print culture context and the pressures and pleasures of Orientalist discourse. Most important, the essay distinguishes Cunninghame Graham from his literary peers, especially his friend Joseph Conrad.

The distribution of fibre bundles in chopped carbon fibre preforms was described using a stochastic model, as a function of geometrical and preform processing parameters. Local permeability distributions were generated from this, based on Gebart’s model for the permeability of aligned fibres. Resin injection simulations were evaluated statistically based on these permeability distributions. Results indicate that, as expected, the macroscopic preform permeability decreases with increasing superficial density (i.e. the average fibre volume fraction). It is independent of the tow filament count and the fibre length within the ranges investigated. The observed coefficient of variation of the permeability increases with increasing filament count, fibre bundle length and superficial density. At constant preform superficial density, variation of the preform processing parameters (fibre spray path offset, spray tool elevation, and spray path pattern) affects the fibre bundle distribution and thus the flow front patterns. Different combinations of the processing parameters have different effects on the mould fill times.

This article describes a method of using both picture frame (PF) and bias extension (BE) tests together to characterize accurately the trellis shearing resistance of engineering fabrics under low in-plane tension conditions. Automated image analysis software has been developed to reduce the amount of laborious manual analysis required to interpret BE data accurately. Normalization methods for both PF and BE tests on rate-independent compressible fabrics are presented. Normalization of PF test results is relatively straightforward while normalization of BE test results for direct comparison with PF data is more complicated. The normalization method uses a number of simple assumptions to account for the nonuniform shear strain field induced across BE samples during testing. Normalized results from BE tests on samples of different aspect ratios are compared and provide validation of the theory.