Documente Academic
Documente Profesional
Documente Cultură
www.elsevier.com/locate/compscitech
Received 23 May 2000; received in revised form 11 November 2000; accepted 13 March 2001
Abstract
Eect of laminate conguration on the impact behaviour of dierent polymer-matrix composites subjected to a transverse central
low-velocity point impact load has been studied. For this a 3D transient nite-element analysis code using a modied Hertz law has
been used. Quadratic failure criteria have been used to predict in-plane and interlaminar failure initiation. The studies have been
carried out with plate dimensions of 150 mm150 mm6 mm with a simply supported boundary condition. For these studies, an
incident impact velocity of 3 m/s and an impactor mass of 50 g have been used. Studies have been carried out on dierent mixed
composites, cross-ply laminates, woven-fabric composites and 3D composites. It is observed that mixing of unidirectional and
woven-fabric layers leads to the reduction of the failure function. # 2001 Published by Elsevier Science Ltd. All rights reserved.
Keywords: A. Polymer-matrix composites; A. Textile composites; B. Impact behaviour; C. Failure criterion; C. Finite-element analysis; Mixed
composites
properties. Such composites have reduced in-plane Laminate conguration, specically the reinforcing
mechanical properties because of bre undulation. arrangement can have a signicant eect on the impact
Three-dimensional composites are emerging as the behaviour of polymer-matrix composites. To look into
main structural materials where multidirectional this aspect, studies have been carried out on four types
mechanical and thermal stresses are found. Because of of composites: laminated composites made of UD pre-
the through-thickness reinforcement, 3D composites are preg tapes, 2D WF composites, mixed composites made
more delamination resistant. The reinforcement limits using UD prepreg tapes and 2D WF layers and 3D
damage initiation and extension after impact within the woven composites.
composite. They also display better resistance to crack In this paper, the impact behaviour of polymer-matrix
propagation and less notch sensitivity as compared to composite plates simply supported on all four sides and
laminated composites. The spatial framework of the subjected to a transverse central low velocity point
reinforcement results in a sharp increase in load bearing impact load is studied. Prediction of in-plane failure of
capacity of the material in thick-walled structures, layers in the form of matrix cracking/lamina splitting
especially in the zones of application of concentrated and delaminations are the primary objectives of the
loads, cut-outs, ribs under unsteady forces and tempera- present study. The stress state in the composite plate has
ture eects. For the same bre volume fractions, 3D been evaluated using 3D transient nite-element analy-
orthogonally woven composites show a signicant increase sis. Quadratic failure criteria have been used to deter-
in through-thickness properties without a comparable mine the failure function indicating the induced stress
reduction in in-plane properties. The 3D composite struc- level with respect to in-plane failure and interlaminar
tures are manufactured by resin transfer moulding of dry failure.
bre preforms. Hence, complex shapes can be formed.
Mixed composites are formed by a rational combina-
tion of UD layers and 2D textile layers. A unique com- 2. Analytical formulation
bination of higher in-plane properties of UD layers and
balanced in-plane properties of 2D textile layers is made In the impact analysis of the composite plates, it is
use of for improving the overall performance of mixed assumed that the material of each layer is linearly elastic
composites. In hybrid composites, two or more reinfor- and obeys the generalised Hookes law. It is based on
cing materials are used to improve the overall char- small deection theory. For the calculation of the contact
acteristics of the material. force caused due to impact on the plate, the impactor is
Considering the importance of impact behaviour of modelled as an isotropic elastic body of spherical shape
polymer-matrix composites, extensive research has been and the target as a plane orthotropic surface. The impac-
carried out during the last few years on this subject. tor is assumed to be rigid and of higher stiness compared
Review articles on the impact behaviour of polymer- to the target in the direction of impact. The stress eld in
matrix composites covering contact laws, impact dynam- the composite plate was analysed assuming quasi-static
ics, stress analysis, damage initiation and propagation, loading [1921].
damage tolerance and improvements in damage resistance The in-house nite-element analysis (FEA) code was
and tolerance are available in literature [16]. Many used for the damage initiation studies [19,20]. This code
research publications are available on impact behaviour is developed using modied Hertz contact law and a 3D
of polymer-matrix composites covering specic aspects. transient nite element analysis. The overall response of
Some of the typical studies are given in Refs. [718]. the plate was obtained based on the response of the
Matrix deformation and microcracking, interfacial plate due to a unit impulse of time duration t [22].
debonding, lamina splitting, delamination, bre break- Eight-noded brick elements with three degrees of free-
age and bre pullout are the possible modes of failure in dom per node were used for the analysis. The analysis of
polymer-matrix composites subjected to impact loading. the dynamics of impact of the composite plate by a rigid
Even though, bre breakage is the ultimate failure impactor was performed to determine the contact force
mode, the macro damage would initiate in the form of history and the response of the plate.
matrix cracking/lamina splitting and delamination.
Additionally, contact induced crushing of upper layers
is another possibility. Composites with higher damage 3. Laminate congurations studied
resistance are necessary for their eective use.
Generally, it is observed that the woven-fabric (WF) Impact behaviour of 14 types of polymer-matrix
composites have higher impact damage resistance than composites with dierent laminate congurations has
the cross-ply (CP) counterparts made of UD layers [18 been studied. Specically, seven types of mixed compo-
20]. However, the eect of laminate conguration on sites with dierent combinations of UD layers and
the impact behaviour of polymer-matrix composites is plain-weave fabric layers have been studied. These seven
not well understood. types of mixed composites are indicated as: MIX-1
N.K. Naik, S. Meduri / Composites Science and Technology 61 (2001) 14291436 1431
Table 2
Mechanical properties of composites with dierent congurations of reinforcement: E-glass/epoxya
Elastic properties
Material E11 (GPa) E22 (GPa) E33 (GPa) G12 (GPa) G13 (GPa) G23 (GPa) 12 13 23 Vf (kg/m3)
UD 30.9 8.3 8.3 2.8 2.8 3.0 0.327 0.327 0.395 0.4 1750
WG02G 20.8 20.8 8.7 3.92 4.2 4.2 0.173 0.279 0.279 0.4 1750
3D-O 20.1 19.9 14.0 4.0 4.0 4.0 0.201 0.296 0.301 0.4 1750
3D-Ob 16.9 19.5 11.7 3.9 3.9 3.9 0.193 0.360 0.362 0.4 1750
UDc 120 7.9 7.9 5.5 5.5 2.97d 0.33 0.33 0.33d 1580
Strength properties
XT (MPa) YT (MPa) ZT (MPa) S12 (MPa) S13 (MPa) S23 (MPa) XC (MPa) YC (MPa) ZC (MPa) Vf
Here 1, 2 and 12 are the induced inplane stress 4.2. Failure functions for dierent laminate
components and XT, YT and S12 are the normal and in- congurations
plane shear strength values.
The damage initiation takes place in the form of Using the in-house FEA code contact force history
matrix cracking/lamina failure when the value of and plate displacement history as a function of impact
inplane failure function, I just exceeds unity. duration have been obtained. Peak contact force, max-
imum displacement, duration of impact and failure
4.1.2. Interlaminar failure function functions have been presented in Table 3 and Figs. 36
Initiation of delamination due to Z and other inter- for all the 14 laminate congurations. In-plane failure
laminar shear stresses is determined using an inter- function history and interlaminar failure function his-
laminar failure function. For this, an interlaminar tory as a function of impact duration have been deter-
failure function based on through-thickness quadratic mined using the quadratic failure criteria as given
failure criterion as proposed by BrewerLagace has earlier. Maximum in-plane failure functions at the bot-
been used [30]. An interlaminar failure function, I is tom surface (I1) and at the top surface (I2) and the
dened as follows: maximum interlaminar failure functions (I9 and I10)
based on average stress criterion at the top two inter-
2 2 2 2 faces have been presented in Table 3. Interlaminar fail-
3t 3c 23 31
I ure function I9 is with respect to the interface at a
Zt Zc S23 S31
distance of 5.25 mm from the bottom surface and I10 is
with respect to top interface.
Here, 3, 23 and 31 are the induced stress compo- The laminates CP0.40GI and CP0.40GII are symmetric
nents and Z, S23 and S31 are the corresponding strength whereas CP0.40GIII and CP0.40GIV are not symmetric.
values. Subscripts t and c indicate tensile and For the impact conditions considered, the failure has
compressive respectively. Superscript - indicates aver- taken place at the bottom layer for CP0.40GIII and
age value over certain area. CP0.40GIV. For all the CP laminates, failure function I1
The delamination initiation takes place when the at the bottom layer is higher than I2 or I10. This indicates
value of interlaminar failure function, I just exceeds that the failure would initiate at the bottom surface
unity. because of the inplane stress components. This is
The point stress criterion predicts higher values of because of lower transverse tensile strength of UD lay-
interlaminar failure function. This is because the calcu- ers. From Table 3, it can be seen that CP0.40GII has
lations are based on the peak stress values. The average lower failure function than for CP0.40GI. The laminate
stress criterion predicts more realistic behaviour. The CP0.40GI has clustered conguration whereas the
interlaminar failure function values based on average laminate CP0.40GII has staggered conguration.
stress criterion are calculated considering an area with It is seen that the damage initiates at the lower layer
the width equal to half the thickness of the plate around for CP laminates because of lower inplane tensile
the point of peak interlaminar failure function based on strength of the layers. One of the ways of delaying the
point stress criterion. Specically, these values were cal- failure initiation would be to introduce balanced WF
culated as follows [27]. layers at weaker regions. Seven types of UD-WF mixed
composites have been studied (Tables 1 and 3) to look
For WF laminates, into possible eect of mixing of layers on impact beha-
" # " # viour of composites. It may be noted that, MIX-1 and
4 2 1 2 1 MIX-6 are symmetric. Compared to the CP laminates,
Iasc Ipsc Ipsc 0:644 Ipsc
5 5 9 the mixed composites have signicantly lower inplane
failure function. But, there is some increase in inter-
laminar failure function for mixed composites. It may
For UD laminates, be noted that for MIX-1, MIX-2 and MIX-4, the max-
" # " # imum in-plane failure function was at a distance of 1.5
7 2 3 2 1 mm from the bottom surface, i.e. in the second and
Iasc Ipsc Ipsc 0:5 Ipsc
10 10 9 third layers from the bottom surface. This is indicated
by L3 in Table 3. The overall failure function, either the
in-plane or the interlaminar, is lower for mixed compo-
Here, Iasc indicates interlaminar failure function based sites compared to CP laminates. This indicates the ben-
on average stress criterion, whereas Ipsc indicates inter- ecial eect of mixing of UD and WF layers to decrease
laminar failure function based on point stress criterion. the failure function during an impact event. But the
For mixed and 3D composites the relation for WF inplane properties of mixed composites may be marginally
composite has been used. lower than those of CP laminates.
1434 N.K. Naik, S. Meduri / Composites Science and Technology 61 (2001) 14291436
Table 3
Impact behaviour of composites: E-glass/epoxy. M=50 g, V0=3 m/s, plate dimensions: LX=150 mm, LY=150 mm and LZ=6.0 mm, simply sup-
ported
Material: Peak Max. Duration of Time to Time to Maximum in-plane Maximum interlaminar
lay-up contact displacement impactc reach fm (ms) reach failure function failure function (asc)
force fm (N) m (mm) tf (ms) m (ms)
I1 (bottom) I2 (top) I9 I10
CP0.40GI 870 0.600 1031 (961) 74 511 0.536 0.066 0.075 0.096
CP0.40GII 887 0.636 1018 (955) 74 464 0.436 0.062 0.078 0.099
CP0.40GIII 486a 0.212a 98a (98a) 98a 98a 1.000 0.067a
CP0.40GIV 175a 0.293a 136a (136a) 136a 136a 1.000 0.052a
MIX-1 1115 0.609 854 (898) 790 425 0.026 0.089 0.130 0.179
0.072 (L3)
MIX-2 1004 0.633 1026 (932) 827 449 0.022 0.064 0.129 0.165
0.064 (L3)
MIX-3 1026 0.626 1012 (924) 814 444 0.021 0.064 0.137 0.173
MIX-4 1103 0.611 854 (900) 788 425 0.026 0.089 0.127 0.175
0.076 (L3)
MIX-5 1053 0.628 1023 (926) 815 442 0.021 0.064 0.144 0.184
MIX-6 917 0.634 1012 (942) 827 450 0.390 0.061 0.111 0.140
MIX-7 1130 0.613 860 (901) 787 426 0.027 0.088 0.149 0.184
WG02G 1213 0.602 831 (893) 763 412 0.025 0.094 0.160 0.214
3D-O 1209 0.607 837 (902) 780 415 0.014 0.077 0.062 0.073
3D-Ob 1237 0.625 871 (929) 803 429 0.023 0.146 0.192 0.247
a
Indicates the values at in-plane failure.
b
Considering knockdown factors and crimp reduction.
c
Duration of impact is given at contact force, F=0. The quantity in the parentheses indicates duration of impact at plate displacement, w=0.
In general, it can be seen that MIX-1, MIX-2 and contact force, maximum displacement and failure func-
MIX-4 have lower failure functions compared to tions for dierent mixed composites have been pre-
alternate mixed composites. A typical mixed composite sented in Figs. 3 and 4.
MIX-2 has six UD layers and only two WF layers. Composites MIX-1 and MIX-4 have lower displace-
With this, possible reduction in inplane properties ment compared to alternate mixed composites. The
would be minimum, and there is a signicant reduction peak contact force is higher for mixed composites com-
in failure function compared to CP laminates. Peak pared to CP laminates.
N.K. Naik, S. Meduri / Composites Science and Technology 61 (2001) 14291436 1435
A typical plain weave fabric composite and 3D based on actual geometry [23,24]. For such a cong-
orthogonal woven composites 3D-O and 3D-O+ have uration the overall failure function is higher compared
also been studied. It can be seen that the overall failure to 3D-O, 2D WF composites and mixed composites.
function is the least for 3D-O. This indicates 3D com- Peak contact force, maximum displacement and fail-
posites are more impact resistant compared to the other ure functions for CP laminates, MIX-2, WG02G and
laminates. It may be noted that 3D-O is based on an 3D composites have been presented in Figs. 5 and 6.
ideal conguration [23,24]. In actual 3D composites Overall, it is seen that the 3D orthogonal composite
there can be strand undulation leading to reduction in with an ideal geometry has the least failure function.
mechanical properties. The conguration 3D-O+ is This indicates that such composites are more impact
damage resistant. Considering possible strand undula-
tion during 3D composites fabrication, mixed compo-
sites may be a better choice. Among the mixed
composites, MIX-1, MIX-2 and MIX-4 have lower
overall failure function. Among these, MIX-2 can be a
better choice because of lower number of WF layers and
hence possible lower reduction in in-plane properties.
5. Conclusions
Acknowledgements
References
[7] Joshi SP, Sun CT. Impact induced fracture in a laminated com- [18] Kim JK, Sham ML. Impact and delamination failure of woven-
posite. Journal of Composite Materials 1985;19:5166. fabric composites. Composites Science and Technology
[8] Joshi SP, Sun CT. Impact induced fracture in a quasi-isotropic 2000;60:74561.
laminate. Journal of Composites Technology and Research [19] Naik NK, Chandra sekher Y. Damage in laminated composites
1987;9:406. due to low velocity impact. Journal of Reinforced Plastics and
[9] Wu HT, Springer GS. Measurements of matrix cracking and Composites 1998;17:123263.
delamination caused by impact on composite plates. Journal of [20] Naik NK, Chandra sekher Y, Sailendra M. Damage in woven-
Composite Materials 1988;22:51832. fabric composites subjected to low-velocity impact. Composites
[10] Wu HT, Springer GS. Impact induced stresses, strains, and dela- Science and Technology 2000;60:73144.
minations in composite plates. Journal of Composite Materials [21] Davies GAO, Zhang X, Zhou G, Watson S. Numerical modelling
1988;22:53360. of impact damage. Composites 1994;25:34250.
[11] Sun CT, Manoharan MG. Growth of delamination cracks due to [22] Sankar BV, Sun CT. An ecient numerical algorithm for trans-
bending in a [905/05/905] laminate. Composites Science and verse impact problems. Computers and Structures 1985;20:100912.
Technology 1989;34:36577. [23] Cox BN, Dadkhah MS. The macroscopic elasticity of 3D woven
[12] Choi HY, Chang FK. A model for predicting damage in gra- composites. Journal of Composite Materials 1995;29:785819.
phite/epoxy laminated composites resulting from low-velocity [24] Naik NK, Thuruthimattam BJ. Behavior of 3-D orthogonally
point impact. Journal of Composite Materials 1992;26:2134 woven composites under tensile loading. Journal of Composites
69. Technology and Research 1999;21:15363.
[13] Chou S, Chen HC, Chen HE. Eect of weave structure on [25] Wu HT, Chang FK. Transient dynamic analysis of laminated
mechanical fracture behavior of three-dimensional carbon ber composite plates subjected to transverse impact. Computers and
fabric reinforced epoxy resin composites. Composites Science and Structures 1989;31:45366.
Technology 1992;45:2345. [26] Karas K. Platten untersettlichen stoss. Ingenieur Archiv
[14] Jih CJ, Sun CT. Prediction of delamination in composite lami- 1939;10:23750.
nates subjected to low velocity impact. Journal of Composite [27] Naik NK, Chandra sekher Y, Sailendra M. Polymer matrix
Materials 1993;27:684701. woven fabric composites subjected to low velocity impact: part
[15] Pierson MO, Vaziri R. Analytical solution for low-velocity I damage initiation studies. Journal of Reinforced Plastics and
impact response of composite plates. AIAA Journal Composites 2000;19:91254.
1996;34:163340. [28] Lammerant L, Verpoest I. The interaction between matrix cracks
[16] Ebeling T, Hiltner A, Baer E, Fraser IM, Orton ML. Delamina- and delaminations during quasi-static impact of composites.
tion failure of a woven glass ber composite. Journal of Compo- Composites Science and Technology 1994;51:50516.
site Materials 1997;31:131833. [29] Jones RM. Mechanics of composite materials. Washington (DC):
[17] Ying Y. Analysis of impact threshold energy for carbon bre and Scripta Book Company, 1975.
fabric reinforced composites. Journal of Reinforced Plastics and [30] Brewer JC, Lagace PA. Quadratic stress criterion for initiation of
Composites 1998;17:105675. delamination. Journal of Composite Materials 1988;22:114155.