MODELLING COMPACTION EFFECT ON PERMEABILITY OF 3D CARBON - PDF document

18 TH INTERNATIONAL CONFERENCE ON COMPOSITE MATERIALS MODELLING COMPACTION EFFECT ON PERMEABILITY OF 3D CARBON REINFORCEMENTS XS Zeng*, AC Long, F Gommer, A Endruweit and M. Clifford Division of Materials, Mechanics & Structures, Faculty of Engineering, University of Nottingham, University Park, Nottingham, UK, NG7 2RD * Corresponding author ( xuesen.zeng@nottingham.ac.uk ) Keywords : liquid composite molding, 3D weave, compressibility, permeability, X-ray computed tomography , FE analysis Focus Detector Distance 200mm 1 Introduction o Molybdenum Target (high contrast on low o absorbing material, useful in 20-60kV The relationship between compaction and range) permeability was investigated for 3D woven carbon Mode 1 or 2 (Mode 1 down to 1.2 microns reinforcement. The geometry and structural change o voxelsize, Mode 2 from 0.9 to 1.2 microns) under different levels of compression was captured All power with modus 1 or 2, Voltage= by micro x-ray computed tomography (CT). The o 40keV and Current = 240µA (low tension imaging technique has been used increasingly for and high intensity to increase contrast) composite material characterization [1-3]. The Exposure time 500ms (contrast of advances in CT hardware helps improve image o resolution) resolution and contrast for low absorbing material Exposure average (1500 ms in total) such as carbon fibre [4]. The 3D image data were o Detector skip (500ms) then analyzed to characterize the structural o Min. 2200 projections deformation and variability. Numerical modelling o was performed to predict though-thickness and in- plane permeabiltiy for 3D woven carbon fibre 3 Image analysis of 3D carbon fabrics. This used an automated modelling approach reinforcements from TexGen geometry modelling and Three samples of the same orthogonal weave at discretization, to CFD analysis in Ansys CFX. A different fibre volume fractions (V f ) were scanned as unit cell representing 3D woven fabric geometry was shown in Figure 1. The cross-section images of the generated in an automated manner in TexGen from a reinforcement in dry form and two impregnated set of experimentally determined geometric panels clearly show progressive deformation in fibre parameters. After meshing and application of tows and resin pockets as compaction is applied with periodic boundary conditions, through-thickness and increasing fibre volume fraction. The defects of air in-plane flow were predicted using CFD simulations. voids and cracks in the composite panels are The predictions were compared with experimental observed due to imperfections in the manufacturing permeability measurements. process, which do not affect the validity of geometric observations. 2 Image acquisition by computed tomography The µCT data contain thousands of grey scale image slices which are to be re-stacked in warp, weft X-ray µCT was performed on a Pheonix and through thickness directions for geometry Nanotom X-ray scanner (GE Sensing & Inspection measurement. In order to acquire data from a large Technologies GmbH). In order to achieve high number of images, an in-house MatLab code was quality images in resolution and contrast for carbon developed to process CT images and take fibre composites, the following configurations were measurements. The target area for permeability used: study is resin flow channels including gaps between Small sample (5 x 5 x 20 mm which is o yarns and compaction plates. As shown in Figure 2, slightly larger than unit cell size of 3D the code first identifies the flow channel boundary woven reinforcement) by searching grey scale contrast between fibre tows Focus Object Distance as small as possible o

and resin pockets. Each flow region is labeled so V f = 0.50; middle, composite panel V f =0.55; bottom, that the continuity of each flow channel in 3D space highly compacted composite panel, V f = 0.64. ሺ௫ିఏሻఙ √ ଶగ ݁ ି ሾౢ౤ ( ೣషഇ ) షഋሿమ can be traced. Measurement of area, perimeter and ଵ ݂ ( ݔ )= ǡݔ൐ Ͳ మ഑మ (1) centre coordinates is output in a tabulated data file. The image is then converted to black and white The three parameter form of the lognormal binary format. The image sequence is used to render distribution function has parameters σ the shape flow channels in 3D as shown in Figure 3. The warp parameter, µ the median (a scale parameter) and flow channels are disconnected where binder yarns θ as location (shift) parameter. pass through them. The channel is narrower in the middle of each segment, indicating warp yarn bulges in the middle due to compaction. Each channel shares similarity differences exist between channels. The variability in 3D woven reinforcement is thought to comprise periodic changes such as yarn cross-section variation and stochastic deviation. The measurements from the image analysis allow both systematic variation and deviation probability to be characterised. An example is given for area measurement in Figures 4 and 5. Similar characterization is applied to flow path and cross- section aspect ratio. A lognormal distribution function is used to describe the stochastic behaviour of the flow channel area. Fig. 2. Image segmentation of flow channel in 3D carbon reinforcement. Upper, flow channel regions are identified and labeled in original CT image; lower, converted binary image with flow channels. Fig.1. CT images of 3D orthogonal carbon Fig. 3. 3D render of isolated flow channels in warp direction of carbon composites V f =0.64. reinforcement. Top, dry textile with no compaction,

After generating a uni unit cell geometry model, TexGen starts generates a a voxel mesh with uniform hexahedral elements for the the whole unit cell domain. The elements are groupe uped by a local sampling technique as either in flu fluid domain or individual yarn volumes. Each elem ement is identified by its greyscale value. Greyscale ale value 0 is assigned to elements at flow area and nd a value range from 1 to 255 is for sequential separat rate yarns. The voxel mesh data fr from TexGen are exported to ANSYS CFX pre-proce ocessor. Periodic boundary Fig.4. Averaged flow cross-section ion area over a conditions are set for the v he voxel mesh model in both periodic channel length th weft and warp directions ns to represent the infinite 3-Parameter Lognormal size of fabric. A pressure re drop is applied to weft, 35 Loc -2.405 warp and top/bottom bounda undaries to predict weft in- Scale 0.035 Thresh -0.09 30 N 1138 plane, warp in-plane a and through thickness 25 permeability. Air at 25 C and and 1 atm (dry) is selected as the medium for steady st state flow through fabric. Frequency 20 15 5 Permeability prediction ctions 10 The flow simulation on is performed for the 5 carbon reinforcement wi with V f = 0.55 and 0.64 respectively. The fabric permeability 0 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.0 0.020 Area deviation measurements labelled ed as Fabric 3 were published by Endruweit eit and Long [6]. It is Fig. 5. Probability distribution of ar area deviation from the average value. ue. noted that the measured sured in-plane permeability values are given in the pr he principle material axes, 4 CFD modeling for permeability lity prediction which is at an angle to wa warp and weft directions. The geometric model of 3D carbon bon reinforcement The permeability in warp arp and weft direction is is generated by using TexGen whic hich is an open calculated by the conver ersion from the principle source software package for textile mo modelling. It has and K 2 , and the angle α axial permeability K 1 and been developed at the University o of Nottingham. between the principle ax axis and weft yarn as in The modelling algorithm is detailed elsewhere [5]. Equation 2. The graphical user interface (GUI) pr ) provides access to major TexGen functions. Alt Alternatively an application programming interface e (API) enables     2 2 K K cos K sin (2) weft 1 2 Python scripting language. TexGen en is capable to      model a variety of 3D textiles orie riented for finite 2 2 K K sin K cos warp 1 2 element analysis. The voxel mesh appr approach enables quick The primary definition of any text textile in TexGen adaption of the TexGe Gen geometry model to is to use a centreline describing yar arn paths in 3D study the influence of of periodic geometry space with superimposed cross sectio tions. In order to variations on permeabili bility. The current study establish a predictive model, the geom ometry modelling looks at changes of yarn rn cross-section and yarn is not to reproduce all details as obser bserved from µCT path within a unit cell de ll definition. In Figure 6 analysis. Instead, general rules are dra drawn from µCT left, a TexGen geometry ry model is shown for the analysis and then applied in TexGen t n to describe key orthogonal weave with ith V f = 0.55 under variations of yarn path and yarn rn cross-sections compaction. Local geometry geom variations are according to the level of compaction on. The aim is to considered. Binder yarn rn is the dominant feature determine what information should be be included in an governing flow channe nnels. The binder cross- authentic CAD model for analysis. . In this study, permeability is examined through peri periodic geometry section is squashed on the the surface layers of the variations. orthogonal weave, while hile weft yarns sink in at