F inite element modelling of progressive damage in non-crimp 3D orthogonal weave and plain weave E -glass composites Stepan V. LOMOV, Dmitry S. IVANOV, Ignaas VERPOEST Department MTM, Katholieke Universiteit Leuven, Belgium Alexander E. BOGDANOVICH 3Tex Inc, USA Kenta HAMADA, Tetsusei KURASHIKI, Masaru ZAKO Osaka University, Japan Mehmet KARAHAN Uludag University, Turkey 1 3D Textiles Greenville 2009
C ontents 1. Introduction: Materials and experimental results 2. FE models and progressive damage model 3. Results of FE analysis and comparison with experiments 4. Conclusions 2 3D Textiles Greenville 2009
1. Introduction • meso-FE analysis of textile composites • Materials: 3D non-crimp fabrics and plain weave laminate • Experimental results 2. FE models and progressive damage model 3. Results of FE analysis and comparison with experiments 4. Conclusions 3 3D Textiles Greenville 2009
meso-F E : R oad map Geometric modeller Geometry corrector Meshing Assign material N+2 N+1 N properties Boundary conditions FE solver, postprocessor Homogenisation Damage analysis 4 3D Textiles Greenville 2009
WiseTex–MeshTex/S A C O M State-of-the-art numerical W iseTex Geometric modeller tool for preparation of FE models and FE analysis of Geometry corrector textile composites on meso-structural level Meshing M esh Tex Assign material properties Boundary conditions FE solver, postprocessor SACOM , Homogenisation Visu a l SACO M Damage analysis 5 3D Textiles Greenville 2009
Internal structure of 3D and plain weave composites 1. Almost straight yarns 2. Slight crimp of the fill caused by compaction in VARTM Plain weave laminate Crimped warp/weft, nested plies 6 3D Textiles Greenville 2009
P arameters of 3D and plain woven fabric 3D – 96 (oz/sq.yrd) 2D – 24 (oz/sq.yrd) 4 plies: 0° /90° /90° /0° Fabric and composite plate 1 ply Fabric and composite plate 4 ply Areal density, g/m2 3255 Areal density, g/m2 3260 Thickness, mm 2.6 Thickness, mm 2.45 Ends (straight) per cm per layer 2.76 Ends per cm 5.08 Picks per cm 2.64 Picks per cm 6.19 Z-yarns per cm 2.76 VF, % 48.9 VF, % 52.4 Yarns tex Yarns tex Warp Warp and weft 2275 layer 1,3 2275 layer 2 1100 Z-yarns 276 Fill (double yarns) layer 1,4 1470 layer 2,3 1470 Warp : Fill : Z = 49% : 48% : 2% 7 3D Textiles Greenville 2009
Q uasi static tension with acoustic emission and strain mapping 500 1.E+10 2D-24 1.E+09 400 1.E+08 1.E+07 stress, MPa 300 1.E+06 AE 1.E+05 200 1.E+04 stress-strain 1.E+03 100 AE events 1.E+02 AE cumulative 0 1.E+01 0 0.5 1 1.5 2 2.5 3 3.5 strain, % � 1 � 2 1.E+04 1.E+09 cumulative AE energy energy 1.E+08 energy of 5.E+03 events 1.E+07 b AE energy 0.E+00 1.E+06 strain, % � 1 Displacement-controlled 0 0.2 0.4 0.6 0.8 1 1.E+05 1.E+08 tension (Instron) � min � 2 1.E+04 AE energy Acoustic emission (Vallen) 1.E+03 5.E+07 1.E+02 Strain-mapping (LIMESS) 0 0.2 0.4 0.6 0.8 1 c 0.E+00 strain, % a strain, % 0 0.2 0.4 0.6 0.8 1 8 3D Textiles Greenville 2009
E xperimental results 600 Tension in warp/fill direction, normalised @VF=50% 500 3D-96, warp 400 3D-96, fill stress, MPa 2D-24 300 3D non-crimp composites: 200 higher strength 100 0 0 0.5 1 1.5 2 2.5 3 3.5 strain, % 3D non-crimp composites: damage initiation delayed 9 3D Textiles Greenville 2009
D amage development 10 3D Textiles Greenville 2009
1. Introduction 2. FE models and progressive damage model • Geometrical models • Meshing • Progressive damage model 3. Results of FE analysis and comparison with experiments 4. Conclusions 11 3D Textiles Greenville 2009
3D finite element model of 2D woven composite One ply Clearance between yarns 0.005 mm Resin layer on the surface 0.005 mm W iseTex VF 47.2% (WiseTex: 52.0%) Total elements 8512 Penetrating nodes corrected 1613 Max aspect ratio 1337 Max aspect ratio in yarns 64 M esh Tex Correct representation of measurable parameters: • areal density • thickness • overall fibre volume fraction Mesh in the yarns • ends/picks count • yarn dimensions Simplifications: • modelling of one ply with periodic BC in the thickness direction Full mesh • unbalanced ply vs balanced laminate 12 3D Textiles Greenville 2009
3D finite element model of 3D woven composite W iseTex Clearance between yarns 0.005 mm Resin layer on the surface 0.005 mm VF 43.7% (WiseTex: 48.9%) Total elements 20768 Penetrating nodes corrected 3000 Max aspect ratio 469 M esh Tex Max aspect ratio in yarns 60 Correct representation of measurable parameters: • areal density • thickness • overall fibre volume fraction Mesh in the yarns • ends/picks count • yarn dimensions Simplifications: • elliptical shape of yarn cross-sections • constant dimensions of Z-yarns • VF inside yarns up to 90% (Z-yarns) Full mesh 13 3D Textiles Greenville 2009
D amage model (built-in in S A C O M) –1 Damage initiation: Hoffmann Definition of the damage mode Z 2 2 2 ( ) ( ) ( ) F C C C � � � � � � 1 � T � Z 2 � Z � L 3 � L � T T 2 2 2 C C C C C C L � � � � � � 4 � 5 � 6 � 7 � 8 � 9 � L T Z TZ ZL LT 1 1 1 1 � � � C � � � � � � 1 t c t c t c � � 2 F F F F F F � T T Z Z L L � � � 1 1 1 1 � � � C � � � � � 2 � t c t c t c � 2 F F F F F F � � Z Z L L T T � � 1 1 1 1 � � � C � � � � � � 3 t c t c t c � � 2 F F F F F F � � L L T T Z Z � � 1 1 1 1 1 1 � , , C C C � � � � � � 4 5 6 t c t c t c F F F F F F � L L T T Z Z � 2 2 2 � 1 1 1 � � � � � � C , C , C � � � � � � � � � � 7 8 9 � s � � s � � s � F F F � � TZ � � ZL � � LT � � � 14 3D Textiles Greenville 2009
D amage model (built-in in S A C O M) –2 15 3D Textiles Greenville 2009
U D strength parameters Corrected (L) for 75% and UD [1], VF=60% [6], VF=55% Hybon’ data accepted for calculations L, tensile 1020 1080 1380 1725 L, compression 620 620 620 T, tensile 40 39 40 T, compr 140 128 140 LT 70 89 70 TZ 70 Matrix tensile 76 Compression 112 shear 88 [1] "Composites Engineering Handbook" (P.K. Mallick, Ed.), Marcel Dekker, Inc., New York, 1997 (Table 1) [2] "Engineering Mechanics of Composite Materials" by I.M. Daniel and O. Ishai, Oxford University Press, New York - Oxford, 1994 (Table 2.6) 16 3D Textiles Greenville 2009
1. Introduction 2. FE models and progressive damage model 3. Results of FE analysis and comparison with experiments • Tensile diagram • Strength • Damage initiation • Damage propagation and damage modes 4. Conclusions 17 3D Textiles Greenville 2009
E lastic properties: correct! 18 3D Textiles Greenville 2009
vsstrain map. � xx . Tension in warp direction, < � xx > = 0.1 F E % 0.003 0.0025 FE 0.002 FE: z-periodicity 0.0015 Digital image correlation: 0.001 • Free surface • segment 29x29 pix 0.0005 • step 5x5 pix LIMESS • strain calculation 5x5 filter 0 0 500 19 3D Textiles Greenville 2009
vsstrain map. � xx . Tension in warp direction, < � xx > = 0.2 % 3D MO S A IC Free surface Digital image correlation: • segment 29x29 pix • step 5 pix • strain calculation 5x5 filter 20 3D Textiles Greenville 2009
Tensile diagrams, 2D composite • correct modelling of degradation of stiffness • reasonable evaluation of damage initiation threshold • qualitative representation of intensity of damage 21 3D Textiles Greenville 2009
Tensile diagrams, 3D composite • correct modelling of degradation of stiffness • reasonable evaluation of damage initiation threshold • qualitative representation of intensity of damage 22 3D Textiles Greenville 2009
S trength, 2D and 3D composites 2D 3D 1. The calculations for one unit cell do not represent stochastic failure of the full sample. 2. Longitudinal strength value for glass fibres and the strength of the impregnated fibre bundles, assumed in the failure criterion, is based on the fibre manufacturer’s data sheet and may not reflect the actual strength of the fibres after textile processing. 3. The damage model does not include influence of delamination and splitting , evident in 2D composites at the late stage of deformation 23 3D Textiles Greenville 2009
P rogressive damage, 2D composite 24 3D Textiles Greenville 2009
P rogressive damage, 3D composite 25 3D Textiles Greenville 2009
1. Introduction 2. FE models and progressive damage model 3. Results of FE analysis and comparison with experiments 4. Conclusions 26 3D Textiles Greenville 2009
Recommend
More recommend
