(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Static and dynamic strain energy release rates in toughened thermosetting composite laminates"

/• / _-- 



N95- 28288 



STATIC AND DYNAMIC STRAIN ENERGY RELEASE RATES IN 
TOUGHENED THERMOSETTING COMPOSITE LAMINATES 

Douglas S. Cairns" ' 

Hercules Advanced Materials and Systems Company 
(HAMSCO), Magna, UT 

P /o 

Abstract ' 

In this work, the static and dynamic fracture properties of several thermosetting resin based 
composite laminates are presented. Two classes of materials are explored. These are 
homogeneous, thermosetting resins and toughened, multi-phase, thermosetting resin systems. 
Multi-phase resin materials have shown enhancement over homogenous materials with respect to 
damage resistance. The development of new dynamic tests are presented for composite laminates 
based on Width Tapered Double Cantllevered Beam (WTDCB) for Mode I fracture and the End 
Notched Flexure (ENF) specimen. The WTDCB sample was loaded via a low inertia, pneumatic 
cylinder to produce rapid cross-head displacements. A high rate, piezo-electric load cell and an 
accelerometer were mounted on the specimen. A digital oscilloscope was used for data 
acquisition. Typical static and dynamic load versus displacement plots are presented. The ENF 
specimen was impacted in three point bending with an instrumented impact tower. Fracture 
initiation and propagation energies under static and dynamic conditions were determined 
analytically and experimentally. The test results for Mode I fracture are relatively insensitive to strain 
rate effects for the laminates tested in this study. The test results from Mode II fracture indicate that 
the toughened systems provide superior fracture initiation and higher resistance to propagation 
under dynamic conditions. While the static fracture properties of the homogeneous systems may 
be relatively high, the apparent Mode II dynamic critical strain energy release rate drops 
significantly. The results indicate that static Mode II fracture testing is inadequate for determining 
the fracture performance of composite structures subjected to conditions such as low velocity 
impact. A good correlation between the basic Mode II dynamic fracture properties and the 
performance in a combined material/structural Compression After Impact (CAI) test is found. These 
results underscore the Importance of examining rate-dependent behavior for determining the 
longevity of structures manufactured from composite materials. 

Introduction 

With composite materials being used in primary aerospace structures, some basic understanding of fracture 
is necessary. This is especially important to develop methodologies for determining damage resistance and 
damage tolerance of composite structures [1]. "Damage resistance" refers to the ability of a 
material/structure to sustain an "event" without resulting in damage and "damage tolerance" refers to the 
ability of a material/structure to maintain performance with damage present. Composite laminates typically 
have poor, through-the-thickness performance. This makes them especially sensitive to out-of-plane loadings 
such as bending and impact. The goal is to provide an understanding of fracture such that the performance 
of laminated structures subjected to low velocity impact can be obtained. For a preliminary assessment of 
performance, a study of Mode I and Mode II fracture behavior under static and dynamic conditions was 
conducted. These modes of fracture are important for delamination initiation and propagation for thin, 
composite laminates subjected to low velocity impact [1-5]. 



"Research Associate and Manager, Advanced Composites Technology, Magna, UT 



1529 



Toughened Thermosetting Resin Systems 

A relatively new class of thermosetting resin systems for advanced composites is being developed [2-4]. 
These toughened systems are being developed with better damage resistance and damage tolerance to 
improve the longevity of aerospace structures such as commercial aircraft. Bradley has noted that an 
improved process zone in composite laminates is a key for determining interlaminar fracture performance 
[5]. This process zone is the region between plies and is where Interlaminar fracture (delamination) typically 
occurs. Figure 1a is a photograph of a crack (delamination) in the interply of Hercules IM7/3501-6. Notice 
that the crack propagates close to the fibers, along one side of the fiber/matrix in a self-similar manner. 
Some separation of the fracture surface is seen, but the fracture propagates at the ply/matrix interface. This 
is typical of brittle, homogenous, thermosetting resin systems [6]. Figure 1b is a photograph of an interply 
crack in Hercules IM7/8551-7. Notice how the interply crack is not self-similar and propagates through and 
around the different phases. This combination of tough phases and process zone enhancement results in 
laminates which have superior damage resistance during low velocity impact [7]. Consequently, a basic 
study to determine the fracture initiation and propagation in this interply region under static and dynamic 
conditions is necessary for a preliminary understanding of the mechanisms of toughening. 

Mode I Test Method 

For Mode I fracture performance, the Width Tapered Double Cantilevered Beam Test (WTDCB) was used 
[8]. A schematic of the geometry used in this study is shown in Figure 2. This sample exhibits a constant 
fracture load for a given Mode I critical strain energy release (G, c ). From elementary beam theory, for the 
geometry given in Figure 2, the Mode I critical strain energy release rate, G| C , is calculated to be: 



n 12P 1 
K Eh 3 



(f)" 



If the thicknesses of the halves are unequal; *' 



#c £ 



6P 2 ( 1 O '-tf 



J .3 






l*, J Hi) 

where h is the half-thickness of the sample and h 1t h 2 are the thicknesses for unequal thicknesses. Hercules 
uses a nominal specimen which is 152 mm long, 25.4 mm wide, and 3.3 mm thick with a 25 mm precrack. 
The taper ratio (a/b) is equal to 4. 

For static testing, the specimen is simply loaded into a test machine and tested at 2.5 mm/ min. The 
average load during fracture is recorded and used in Equation 1. A schematic for the Mode I dynamic 
testing is shown in Figure 3. In this test setup, a low inertia, pneumatic cylinder is used for actuation. The 
specimen is instrumented with a high-speed, piezo-electric load cell. In addition, an accelerometer is placed 
on the cross-head to monitor accelerations which are numerically-integrated to determine velocities and 
displacements. A typical load-time plot is shown in Figure 4. The total fracture time is approximately 9 ms. 
Notice that, while some deviation in the load is seen during fracture, the load remains relatively constant. 
Consequently, Equation 1 was used for determining Mode I dynamic critical strain energy release rates (G, c ). 
Some typical results, comparing static and dynamic Mode I dynamic critical strain energy release rates, are 
shown in Table 1 . Five coupons from each specimen type were used for the tests. 



1530 



Material Type 


Static 
J/m 2 


(in-lbs/in 2 ) 


IM7/X1 
IM7/X2 
IM7/X4 


394 
256 
530 


(2.25) 
(1.46) 
(3.02) 


IM7/8551-7A 
IM7/8552 


249 
233 


(1.42) 
(1.33) 


IM7/8551-7 
IM7/8551-7C 


552 
381 


(3.15) 
(2.17) 



Table 1. Comparison of Static and Dynamic Mode I Fr acture Properties 

Dynamic G lc (Eq.1) 
J/m 2 (in-lbs/in 2 ) 

312 (1.78) 
286 (1.63) 
500 (2.85) 

202 (1.15) 
240 (1.37) 

607 (3.46) 
456 (2.60) 

In Table 1, the 8552 is representative of a homogeneous resin system and the X1, X2, X4 and 8551 type 
resins are representatives of multiphase systems. The coefficient of variation within a panel is typically 5%- 
6% and the coefficient of variation from panel to panel is typically 10%. With consideration to the variability 
in this test, the data generally show little sensitivity for the rates tested in this study. 

Mode II Test Method 

For these basic studies, the End Notched Flexure (ENF) specimen was chosen as a result of its ease to 
manufacture and test [9-15]. This sample is illustrated in Figure 5. In Figure 5, a beam with a crack in one 
end is loaded in three point bending to produce interlaminar shear at the midplane. The crack length is 
denoted as a and the specimen supported length is 2L with an overall thickness of 2h. Under loading, the 
crack propagates and simple beam theory is used to determine the critical strain energy release rate at 
fracture. From Linear Elastic Fracture Mechanics (LEFM) and simple beam analysis of the ENF, the Mode 
II critical strain energy release rate, G, lc may be calculated as [9-11]: 

G 9P 2 Ca 2 2) 

" c 2w (2L 3 ♦ 3a J ) 

where P is the load at fracture, and C is the overall compliance of the specimen. An additional, small 
correction to account for shearing deformation may be used by replacing C by C [11]: 

c . . C ♦ 12L+ °- 9a 3) 

4whG l3 

where G 13 is the interlaminar shearing stiffness of the specimen. It is noted that, for practical specimen 
geometries, this correction is less than 2% to the compliance from simple beam theory. 

An alternative method for determining the average strain energy release rate during crack propagation may 
be developed by taking data from the load-displacement curve during testing [16]. In this method, P, is the 
load at onset of fracture, and P a is the load at which the fracture is arrested. By assuming a straight line 
between P, and P a the strain energy release rate may be determined as: 

G P > A ° " f « A ' 4) 

* 2w(a a -a) 

where A„A a are the measured displacements at initiation and arrest, and a„ a a are the crack lengths at 
initiation and arrest. Whitney, et. al. [17] have shown this method to be approximately 10% higher than the 
compliance method as in Equation 2 for Mode I fracture and Maikuma, et. al. [15] have shown the area 
method based on energies (which are determined from load versus time data) to be approximately 6% lower 
for dynamic testing in Mode II fracture. These results are well within the scatter of such fracture tests [1 1 , 
17]. 



1531 



To determine the critical strain energy release rates under dynamic conditions, a Dynatup ETI 500 
instrumented impact tower was used. A 5.3 kg mass with an impact velocity of 1 .93 m/sec was used for the 
load introduction (kinetic energy equal to 9.8 joules). The test specimen geometry used for this study was 
149.2 mm long (2L) by 25.4 mm wide (w) by 3.3 mm thick (h) with a notch length of 25.4 mm (a). A thin, 
0.025 mm Teflon film was used for the insert and the crack was propagated to 25.4 mm using Mode I 
propagation. Fiber bridging was not seen at the crack tip in the composite systems tested in this study. 
Hercules IM7 fiber was used for most of the studies. This fiber yields an average laminate flexural stiffness 
of 152 GPa. The average laminate interlamlnar shear stiffness is 5.7 GPa. 

A typical force versus time and displacement versus time is shown in Figure 6. Knowing the mass of the 
impactor and the initial velocity, the displacement of the tup is calculated via numerical integration 
(trapezoidal rule twice) of the directly measured force versus time data. In this figure, the load oscillates up 
to a peak and drops sharply at the point of fracture. The fracture was rather catastrophic. Under impact, 
either the sample did not fracture or the crack propagated in an uncontrolled manner in approximately 1 
msec with very little change in displacement. Unlike constant crosshead displacement loading conditions, 
the fracture propagated over the majority of the length of the sample, beneath the loading tup. This is also 
illustrated in Figure 7 in the load versus displacement plot. While the load oscillates considerably at fracture, 
the displacement remains relatively constant. Equation 2 (with Equation 3) may be used to predict the 
initiation strain energy release rate. Note in Figures 6 and 7, a discontinuity in the behavior is seen during 
fracture. This discontinuity is a result of the flexural wave propagating beneath the load cell. As the crack 
propagates beneath the impactor, the shape of the curve does not change. That is, if frictional effects are 
significant as the crack traverses beneath the loading nose, a significant change in the slope in the 
subsequent behavior would be expected to be seen in the load - displacement curve (Figure 7) as a result 
of coulomb dissipation. Consequently, frictional effects as the crack propagates beneath the loading tup 
were assumed to be no more severe than static testing. However, additional studies may be warranted 
based on some of the unexpected results presented below. 

Using Bernoulli-Euler beam theory, the fractured specimen compliance can be determined as: 

c = -161* + 36I 2 a - 181a 2 + 3a» 
iEwh 3 

In Equation 5, the crack is assumed to traverse beneath the impactor. This equation is used in a combined 
analytical-experimental approach to determine the load at crack arrest. By measuring the crack length and 
using Equation 5, the arrest load is predicted and plotted in Figures 6 and 7, denoted as P a . This 
corresponds reasonably well with the average dynamic load at fracture arrest. Consequently, the work of 
fracture , W, to utilize for average fracture energies during dynamic propagation is: 

A - P 

W = f P dA - -± A 6) 

{ 2 • 

where the integral is the area under the experimental force versus displacement curve and P a is determined 
via Equation 5. Some vibratory behavior is present in Figure 6, and Equation 6 represents the average 
dynamic strain energy release from initiation to arrest. This Equation was used in lieu of Equation 4. In 
utilizing this equation, it is implied that dynamic effects in the calculation are negligible. This was shown to 
be true for dynamic tests under similar conditions to the present work by Maikuma, et. al. [15]. 

Experimental Results 

Table 2 is a comparison of Hercules composite systems subjected to static and dynamic Mode II fracture. 
Five (5) coupons were used for each type of test unless noted. The dynamic fracture behavior presented 
in Figures 6 and 7 is a combined material and structural test and is dependent on specimen geometry, 
support conditions, impactor metrics, etc. However, all coupons were tested under the same conditions for 
consistency. 



1532 



All laminates are nominally 34% resin content by weight (nominally 58.8% fiber volume). The X8553, 8551-7 
and X series laminates are multi-phase, process zone enhanced laminates similar to Figure 1b; while the 
3501-6 and 8552 resin type laminates are homogeneous with limited process zones in the interply region, 
similar to Figure 1a [4,5]. The process zone enhanced laminates exhibit higher Mode II toughnesses than 
the homogeneous systems. The X2 resin is a toughened system formulated to have superior hygrothermal 
properties (hot/wet 0° compression performance) but exhibits lower Mode II toughness than other systems. 
Two resin contents were tested In the X2 formulation as noted. The higher resin content X2 system exhibits 
higher Mode II fracture properties. 

In Table 2, there is not necessarily a correlation between static Mode II and dynamic Mode II fracture 
properties. This is especially evident in the case of the homogeneous systems. While the 8552 exhibits the 
poorest static Mode II interlaminar fracture toughness, it has better Mode II dynamic properties. Its damage 
resistance in low velocity impact is superior as well [18]. The AS4/3501-6 exhibits the poorest damage 
resistance in low velocity impact. Notice that it also has the lowest initial dynamic strain energy release rate 
as defined by Equation 2 (with the minor shearing correction suggested in Equation 3). Upon initiation, the 
average fracture resistance as determined by Equations 4 and 6 drops dramatically compared to the other 
materials. The Average Dynamic G Hc (Eq. 6) of the AS4/3501-6 is less than half of the Dynamic G llc (Eqs. 
2,3) and considerably lower than the static Mode II fracture properties. Once fracture initiates, the resistance 
to propagation is lower in the AS4/3501-6 system, resulting in more damage (delamination) under dynamic 
conditions [2]. Similar trends have been noted by other investigators [13,15] for the same 3501-6 resin 
system. The homogeneous resin laminates appear to be much more strain rate sensitive than the toughened 
systems. Hence, static fracture properties may not be applicable for determining the damage resistance, 
damage tolerance and longevity under dynamic conditions. 

Table 2. Comparison of Static and Dynamic Mode II F racture Properties 

Material Type Static G Bc Dynamic G llc (Eq. 2) Average Dynamic G u (Eq. 6) 

J/m 2 (in-lbs/in 2 ) [CV 1 ] J/m 2 (in-lbs/in 2 ) [CV 1 ] J/m 2 (in-lbs/in 2 ) [CV 1 ] 

IM7/8551-7 
IM9/8551-7 
IM7/X8551-7C 

IM7/X1 

IM7/X2 2 

IM7/X2 3 

IM7/X3 

IM7/X4 

AS4/3501-6 
AS6/8552 

1 CV (Coefficient of Variation, % based on five (5) replicates for each test 
2 IM7/X2 at 39% resin content (by weight) 
3 IM7/X2 at 32% resin content (by weight) 
4 average of two samples 

Compression After Impact 

Compression After Impact (CAI) testing is often used for early screening assessment of performance of 
materials [2-4,7,18,19]. In this test, a quasi-isotropic panel (102 mm x 152 mm X 4.5 mm) thick is impacted 
with 667 N-m/m(thickness) impact energy at low velocity (approximately 2.5 m/s) with a 12.7 mm diameter 
tup. This test is a combined damage resistance and damage tolerance test wherein damage is introduced 
during the impact event and subsequently tested in an end loaded compression fixture with semi-clamped 

1533 



1867 


(10.6) 


[5.6] 


2730 


(15.5) 


[39] 


2272 


(12.9) 


[8.3] 


1691 


(9.6) 


[7.5] 


2184 


(12.4) 


[7.2] 


1832 


(10.4) 


[10.5] 


1374 


(7.8) 


[2.6] 


2360 


(13.4) 


NA 4 


1744 


(10.2) 


[12.7] 


1268 


(7.2) 


[5.9] 


2202 


(12.5) 


[4.7] 


2184 


(12.4) 


[3.5] 


1426 


(8.1) 


[3.5] 


1726 


(9.8) 


[3.5] 


1920 


(10.9) 


[2.4] 


1268 


(7.2) 


[5.9] 


1691 


(9.6) 


[8.4] 


1761 


(10.0) 


[6.3] 


1374 


(7.8) 


[7.8] 


1814 


(10.3) 


[19.7] 


1462 


(8.3) 


[27.1] 


1532 


(8.7) 


[5.6] 


2078 


(11.8) 


[6.9] 


2078 


(11-8) 


[4.3] 


740 


(4.2) 


[7.1] 


916 


(5.2) 


[5.3] 


463 


(2.6) 


[7.0] 


617 


(3.5) 


[6.1] 


1532 


(8.7) 


NA 4 


1391 


(7.9) 


NA 4 



ends. A schematic of the CAI test setup is shown in Figure 8. A plot of CAI performance versus the Average 
Dynamic G llc presented in Table 2 is shown in Figure 9. The high and low data points are shown with error 
bars. 

The CAI test is primarily a damage resistance test. The materials given in Table 1 have similar static damage 
tolerance performance based on open hole compression (OHC) tests and compression tests of laminates 
with similar impact damage [18]. That is, given an equivalent, pre-existing damage state, the materials deliver 
the same static strength. The average OHC performances for the X8553, 8551 -7, X-Series, 3501 -6, and 8552 
based laminates are 300 MPa, 293 MPa, 287 MPa, 300 MPa, and 320 MPa, respectively. However, the 
materials in Table 2 exhibit dramatically different damage resistance performance under dynamic conditions 
such as low velocity Impact. Specifically, the static G llc data presented in Table 2 for the 3501-6 based 
laminate is higher than the 8552 laminate, but this situation is reversed under dynamic conditions. Thus, the 
Average Dynamic G Mc appears to be a leading indicator for a preliminary assessment of the advanced 
composite laminates subjected to lateral impact. It is a measurement of the average strain energy release 
rate for Mode II fracture during delamination formation under dynamic loading conditions. 

Conclusions and Recommendations 

In this study, an assessment of Mode I and Mode II fracture under static and dynamic conditions has been 
developed. Rate had little effect for the rates and materials tested in Mode I in this study. The rates tested 
are typical of the impact duration of low velocity impact [1 ]. Higher testing rates may show a stronger effect. 
In contrast, the Mode II fracture performance of toughened systems is considerably better than 
homogeneous, brittle systems. While the apparent dynamic fracture properties of the systems are higher 
under dynamic conditions, the fracture performance of the toughened multi-phase systems appears to be 
less strain rate sensitive than the homogeneous systems. These results establish the need to examine 
dynamic fracture properties of materials for assessment of damage resistance during low velocity impact. 

The simple, dynamic Mode II fracture test presented herein is somewhat easier to run, and considerably less 
expensive than CAI testing. Also, it has more meaning since it provides a measurement of a basic fracture 
mode, rather than a combined damage resistance, damage tolerance and structural test. The strong trend 
presented in Figure 9 should not be taken as a panacea, as the ultimate goal of increased longevity of 
composite materials is a combined damage resistance and damage tolerance problem. A combination of 
tests under static and dynamic conditions is necessary to determine the longevity of a material and 
structure. 

Future tests should include a wider variety of materials, including thermoplastics, which are known to be 
quite strain rate sensitive [5]. In addition, a study varying the specimen geometry, loading conditions and 
impactor metrics should be conducted to isolate the influence of structural and material parameters of 
fracture performance. 

REFERENCES 

1. Cairns, D.S. and Lagace, PA, "A Consistent Engineering Design Methodology for Composite 
Structures Subjected to Impact", Proceedings of the American Society for Composites: Fifth 
Technical Conference, June, 1990. 

2. Boll, D.J., Bascom, W.D., Weidner, J.C. and Murri, W.J., "A Microscopy Study of Impact Damage 
of Epoxy-Matrix Carbon-Fiber Composites," Journal of Materials Science, Volume 21, 1986. 

3. Cairns, Douglas S., "Prediction of Fracture Toughness of Multi-Phase Materials", Proceedings of 
the AIAA/ASME/ASCE/AHS/ASC 31st Structures, Structural Dynamics and Materials 
Conference, April, 1990. 

4. Cairns, D.S., "Mechanisms of Fracture and Toughening in Multi-Phase Materials", Proceedings of 
the 22nd International SAMPE Technical Conference, Volume 22, November 6-8, 1990. 

1534 



5 Bradley, W.L, The Role of Matrix Properties on the Toughness of Thermoplastic Composites," 

Prepared for Thermoplastic Composite Materials, Series Editor - R. Byron Pipes, Volume Editor - 
Leif A. Carlsson, Elsevier Science Publishers, 1989. 

6. Lagace, P.A. , "Static Tensile Fracture of Graphite/Epoxy," TELAC Report 182-4, Massachusetts 
Institute of Technology, Cambridge, Massachusetts, 1982. 

7. Dost, E.F., llcewicz, LB., and Gosse, J.H., "Sublaminate Stability Based Modeling of Impact- 
Damaged Composite Laminates," Proceedings of the American Society for Composites: Third 
Technical Conference, September, 1988. 

8. Bascom, W.D. and Hunston, D.L, "The Fracture of Epoxy and Elastomer-Modified Epoxy Polymers," 
Treatise on Adhesion and Adhesives, Volume 6, Edited by Patrick, Robert L, Marcel Dekker, Inc., 
New York, 1989. 

9. Russell, A.J. and Street, K.N., "Moisture and Temperature Effects on the Mixed-Mode Delamination 
Fracture of Unidirectional Graphite/Epoxy, Delamination and Debonding of Materials, ASTM STP 
876 (1985), American Society for Testing and Materials, Philadelphia, PA, p. 349. 

10. Carlsson, LA. and Pipes, R.B., Experimental Characterization of Advanced Composite Materials, 
Prentice-Hall, Englewood Cliffs, New Jersey, 1987. 

11. Carlsson, LA. and Gillespie, J.W., Jr., "On the Design and Analysis of the End Notched Flexure 
(ENF) Specimen for Mode II Testing," Journal of Composite Materials, vol. 20, 1986. 

12. Smiley, A.J. and Pipes, R.B., "Rate Effects on Mode I Interlaminar Fracture Toughness in Composite 
Materials," Journal of Composite Materials, vol. 21, 1987, pp. 671-687. 

13. Smiley, A. J. and Pipes, R.B., "Rate Sensitivity of Mode II Interlaminar Fracture Toughness in 
Graphite/Epoxy and Graphite/PEEK Composite Materials," Composites Science and Technology, 
vol. 29, 1987, pp. 1-15. 

14. Friedrich, K., Walter, R., Carlsson, LA., Smiley, A.J., and Gillespie, J.W., Jr., "Mechanisms for rate 
effects on interlaminar fracture toughness of carbon/epoxy and carbon/PEEK composites," Journal 
of Materials Science, vol. 24, 1989, pp. 3387-3398. 

15. Maikuma, H., Gilespie, J.W., Jr., and Wilkins, D.J., "Mode II Interlaminar Fracture of the Center Notch 
Flexural Specimen under Impact Loading," Journal of Composite Materials, vol. 24, 1990, pp. 125- 
149. 

16. Wu, EM., "Fracture Mechanics of Anisotropic Plates," in Composite Materials Workshop, Tsai, 
S.W., Halpin, J.C. and Pagano, N.J., editors, Technomic Publishing Co., Westport, CT, 1968, pp. 20- 
43. 

17. Whitney, J.M., Browning, C.E. and Hoogsteden, W. "A Double Cantilever Beam Test for 
Characterizing Mode I Delamination of Composites," Journal of Reinforced Plastics and 
Composites, vol. 1, 1982, p. 297. 

18. Hercules Prepreg Tape Materials Characterization Data Package, Fibers: AS4, IM6, IM7 & IM8, 
Resins: 8551-7, 8551 -7A, 8552, 3501-6, Hercules Composite Products Group, Magna, Utah, 
November, 1988. 

19. Dost, E.F., llcewicz, Avery, W.B., and Coxan, B.R., The Effects of Stacking Sequence on Impact 
Damage Resistance and Residual Strength for Quasi-lsotropic Laminates," Proceedings of the 
American Society for Composites: Fifth Technical Conference, June, 1990. 

1535 




a) AS4/3501-6 (self-similar crack at ply/matrix interface) 




b) IM7/8551-7 (non self-similar, tortuous crack in interply process zone) 
Figure 1. Crack Propagation in Hercules a) AS4/3501-6 and b) IM7/8551-7 



1536 



l£9l 



Load, N 




o 



-n 

<5' 

c 

s 

rn 

3 
Q. 

z 

O 

o 

IT 
CD 
Q. 



CD 
X 

c 

CD 



m 



CD 

o 
3 

CD 




Id' 

C 
— t 
CD 



Q. 

5 



Q3 TJ 



^ 


CD 

—1 




CD 




Q. 




O 


n 


O 


O 


rr 


UJ 


CD 


O O 


CD 


u> 


O 


3 


CD 


CD 

< 


■3 


CD 




CD 




Q. 



cq 

a 
—% 

CD 
CO 



CD 

CO 

CD 

<-* 

c 

T3 



o 

1 

n 
O 



CD 



3 
CO 



O 



ro 





/£3? 



8CSI 



Tl 
Id' 

C 
-^ 
CD 

00 
CO 

o 

3- 

CO 

3 

CO 

o' 

o 

o 

o 
3 
■o 

CD 
CO 
CO 

o' 

3 
> 

CD 

-^ 

3 
-a 
o> 
o 

O 
> 



CD 
CO- 
CO 
CD 




(Q 

c 

CD 



o 

CO 

a. 

< 

CD 
CO 

c 

CO 



3 

CD 



a 
>< 

3 
CO 

3 
o 

8. 

CD 



CO 

O 

»-» 

c 

CD 

H 

CD 
CO 

5' 
to 



Load, N 

500 1000 1500 2000 
o -}-=; — ■ 1 l 1 1 



3 

CD 

3 
CO 







r 



I: 



r/i 
■H 



15 
Deflection, mm 



— i — 
20 



:;5 



<5" 

c 
-t 

CD 
P 

o 
o 

3 

■D 
-^ 
CD 
CO 
CO 

o' 

3 

> 

CD 

3 

T3 
CO 

O 

r-* 

TJ 

CD 

3- 
O 

3 

CO 
3 
O 
CD 

< 

CD 

-i 
CO 

c 

CO 

> 

< 

CD 

-i 
CO 

co 

CD 

O 

3 
CO 

3 
o 

O 



Compression After Impact (CM), MPa 



-* -* ro 



> 
< 

ro 

3 

CO 
(0 

D 
>< 

CO 

3 
o' 
o 



! ' I 



H 






-n 

-^ 

CD 
^1 



O 

a 
< 

CD 

-i 
CO 

c 
co 

a 

CD 

cb~ 
o 

6" 

3 



Load, N 



D 
•< 

3 
CO 

3 
o 



8. 



CO 

o 

CD 



CD 
CO 



3 



O 
ro 

ro~ 
o 

5' 

3 

3 



500 


1000 1500 


2000 




■in 

■er. 






c 


' — >_ 






< 




- 1 -, 




r-n 


-a 




_TJ 


i — i 



O 




ni 
.a . 
c 



n. 

r.-'i 



/s&