Key Engineering Materials Vols. 611-612

Abstract: The aim of this work is to focus on the Stokes-Darcy coupled problem in order to propose a robust monolithic approach to simulate composite manufacturing process based on liquid resin infusion. The computational domain can be divided into two non-miscible sub-domains: a purely fluid domain and a porous medium. In the purely fluid domain, the fluid flows according to the Stokes' equations, while the fluid flows into the preforms according to the Darcy's equations. Specific conditions have to be considered on the fluid/porous medium interface. Under the effect of a mechanical pressure applied on the high deformable preform/resin stacking, the resin flows and infuses through the preform which permeability is very low, down to 10^-15 m². Flows are solved using finite element method stabilized with a sub-grid scale stabilization technique (ASGS). A special attention is paid to the interface conditions, namely normal stress and velocity continuity and tangential velocity constraint similar to a Beaver-Joseph-Saffman’s condition. The originality of the model consists in using one single mesh to represents the Stokes and the Darcy sub-domains (monolithic approach). A level set context is used to represent Stokes-Darcy interface and to capture the moving flow front. This monolithic approach is now perfectly robust and leads to perform complex shapes for manufacturing process by resin infusion.

Abstract: During the pre-forming stage of the RTM process, large deformations can occur, especially for double-curved shapes. Knowing the mechanical behaviour and the actual geometry of fibrous reinforcements at the mesoscopic scale is of great importance for several applications like permeability evaluations. As such, forming modeling is particularly demanding on the quality of geometric modeling and of the mesh associated. Indeed, analysis of the internal structure of materials in general, and woven materials especially, recently led to major advances. X-ray Micro Tomography (XRMT or μCT) allows detailed and accurate 3D observations inside the sample, which is not possible with the standard microscopy techniques restrained to surface observations. It distinguishes the yarns and even the fibers that define the directions of anisotropy of the material. A FE model is generated from the processed tomography images. It has been chosen in this study to use hypoelasticity behaviour law. Indeed, the yarns are submitted to large deformations, so that the orientation of the material is significantly modified and the fiber direction has to be strictly followed in order to fulfil the principle of objectivity. A way to retrieve the neutral composite reinforcement axis by skeletonization is proposed in order to know the privileged direction of the yarn and thus implement it in the constitutive law. A comparison between experimental and simulations obtained from μCT and idealized geometry of a transverse compression test on the G0986 is presented.

Abstract: During the manufacturing of fabric-reinforced composite parts using a matched-die compression molding process or liquid composite molding, the fabric may experience local in-plane compressive loads that cause out-of-plane deformations. The waves that result from this outofplane motion can lead to the formation of resin rich pockets (during the infusion stage of a dry fabric) or they may be forced down into a fold by the tooling. Defects such as resin-rich pockets and folds compromise the structural integrity of the formed composite part. Therefore, it is valuable to have a simulation tool that can accurately capture the fabric bending properties and predict the locations where waves or folds are likely to occur as a result of the manufacturing process. The tool can then be used to investigate changes in the forming parameters such that the development of such defects can be mitigated. A hybrid finite element model used with a discrete mesoscopic approach captures the behavior of continuous fiber-reinforced fabrics where the fabric yarn is represented by beam elements and the shear behavior is implemented in shell elements. User-defined material subroutines describe the mechanical behavior of the beams and shells for their respective contributions to the overall fabric behavior. Simulations are used to demonstrate the ability of the modeling approach to predict the amplitude and curvature of out-of-plane waves. The simulation results are compared with experimental data to show the accuracy of the modeling. Additional models are presented to demonstrate the capability of the simulation tool to capture fabric folding.

Abstract: A traxial fabric was investigated for use in composite forming applications. Three stitched layers of fibers, originally oriented at [-60^o/0^o/60^o], comprise the fabric architecture. The mechanical properties of the material are characterized by testing the tensile, shear, and frictional behavior. Conventional shear frame testing methodology assumes that the yarns are originally oriented perpendicular to one another; however, such an assumption is not valid for this particular fabric geometry and must be adjusted. The material behavior is implemented into a discrete mesoscopic finite element model that can predict the response of the material during deformation. Different element types will be investigated to represent the fabric and used to determine the ideal mesh configuration that best captures the fabric behavior. Different modes of deformation will also be studied, and the observed experimental deformation will be compared to the deformation predicted by the finite element model.

Abstract: LCM simulation is computationally expensive because it needs an accurate solution of flowequations during the mold filling process. When simulating large computing times are not compatiblewith standard optimization techniques (for example for locating optimally the injection nozzles)or with process control that in general requires fast decision-makings. In this work, inspired by theconcept of medial axis, we propose a numerical technique that computes numerically approximatedistance fields by invoking computational geometry concepts that can be used for the optimal locationof injection nozzles in infusion processes. On the other hand we also analyze the possibilities thatmodel order reduction offers to fast and accurate solutions of flow models in mold filling processes.

Abstract: Shear thickening fluid (STF) is a non-Newtonian fluid featuring the increased viscosity upon high strain rate applied. Recently, STF-treated aramid fabrics have been researched to enhance the bulletproof efficiency maintaining the lightweight, however their shear properties including tow shearing, which significantly contribute to the bulletproof properties, have not been characterized, in particular under high shear strain rates. In this study, the shear properties of STF-treated aramid fabrics are characterized using a picture frame test. For this purpose, STF is prepared using polyethylene glycol and silica colloids and coated onto aramid fabrics. Varying the shear strain rate by controlling the pulling speed of the picture frame, the effect of STF on the shear properties of the aramid fabric is investigated. Finally, the shear properties of STF-treated aramid fabrics are predicted a multi-scale energy model and compared with the experiments. This prediction is then extended to cover such a high strain-rate situation as the bullet impacts, enabling to determine the mechanism behind the improved bulletproof performance of the STF-treated fabric.

Abstract: In order to decrease CO₂ emissions caused by individual transport, fibre-reinforced composite materials are used to reduce vehicle weight and thus fuel consumption. Low productivity of current processes complicates the introduction of fibre-reinforced materials to high volume series, where weight reduction would have a large impact. This paper presents an experimental preforming environment designed to take into account diverse process requirements of different binder systems and a new tooling concept. By setting different temperature levels on the moulds in areas where the material is drawn in under the blank holders and inside the cavity, prebonding of the fabric plies can be avoided and quality of the preforms is improved. Moreover, cycle time can be reduced as no heating or cooling of the tools is necessary.

Abstract: Mesoscopic simulations of the transverse compression of textile preforms are presented in this paper. They are based on 3D FE models of each yarn in contact with friction with its neighbours. The mesoscopic simulations can be used as virtual compression tests. In addition they determine the internal geometry of the reinforcement after compaction. The internal geometry can be used to compute the permeability of the deformed reinforcement and to calculate the homogenised mechanical properties of the final composite part. A hypoelastic model based on the fibre rotation depicts the mechanical behaviour of the yarn. The compression responses of several layer stacks with parallel or different orientations are computed. The numerical simulations show good agreement when compared to compaction experiments.

Abstract: In textile engineering, simulative methods are used more frequently due to their advantages in material and process design. Finite element models were developed for simulating the mechanical and the draping behaviour of fabrics. For large deformation analysis of textile forming, macro mechanical models are employed that use continuum mechanical approaches for matters of reduced computation time. The material data that is required as model input, such as tension and shear properties, can either be obtained by experimental or virtual tests. In such virtual tests the deformation behaviour of fabrics can be determined by deforming the structure on the meso level.

Abstract: CFRTP prepreg laminates thermoforming (Continuous Fibre Reinforcements and Thermoplastic Resin) is a fast composite manufacturing process. Furthermore the thermoplastic matrix is favourable to recycling. The development of a thermoforming process is complex and expensive to achieve by trial/error. A simulation approach for thermoforming of multilayer thermoplastic is presented. Each prepreg layer is modelled by semi-discrete shell elements. These elements consider the tension, in-plane shear and bending behaviour of the ply at different temperatures around the fusion point. The contact/friction during the forming process is taken into account using forward increment Lagrange multipliers. A lubricated friction model is implemented between the layers and for ply/tool friction. Thermal and forming simulations are presented and compared to experimental results. The computed shear angles after forming and wrinkles are in good agreement with the thermoforming experiment.