Sie sind auf Seite 1von 16

The impact resistance of composite materials---a review

W.J. CANTWELL* andJ. MORTONt (* Ecole Polytechnique Fdd#rale de Lausanne, Switzerlandl t Virginia Polytechnic Institute and State University, USA)
In this paper the impact response of continuous fibre-reinforced composites is reviewed. An attempt is made to draw together much of the work published in the literature and to identify the fundamental parameters determining the impact resistance of continuous fibre-reinforced composite materials. The effect of varying the properties of the fibre, matrix and interphase are examined as well as the role of target geometry and loading rate on the dynamic response of these materials.

Key words: composite materials; impact testing; impact resistance; residual strength; failure mode; fibre properties; matrix properties; interphase; laminate stacking sequence; geometry; loading rate
Fibre-reinforced composite materials such as carbon, glass and Kevlar ~ fibre-reinforced plastics are finding increasing use in a wide range of both low and high technology engineering applications. Composites offer a number of distinct advantages over more conventional engineering materials such as aluminium and steel. These include higher specific strengths and stiffnesses, superior corrosion resistance as well as improved fatigue properties. Coupled with these improvements in general overall performance is the fact that the cost of manufacturing components from fibre-reinforced plastic is often less than with more conventional metals. This is true not only at the lowcost end of the market occupied by sheet moulding compounds (SMC) but also in the aerospace sector where complex load-bearing shapes can be produced in a limited number of steps, saving time in both joining and assembly. Composite materials do, however, suffer some serious limitations. Perhaps the most significant amongst these is their response to localized impact loading such as that imparted by a dropped tool or runway debris. In recent years many research programmes have been undertaken in an attempt to better understand the impact response of these materials 1-9. The majority of this work has been undertaken on continuous fibre, high performance composites since these materials are finding increasing use in the design of a large number of civil and military aircraft, i.e., in circumstances where the consequences of impact are likely to be most serious. The manner in which composite materials respond to impact loading and dissipate the incident kinetic energy of the projectile is very different to that of metals. For low and intermediate incident energies, metals absorb energy through elastic and plastic deformation 1. Although the latter may cause some permanent structural deformation, its consequences on the loadcarrying capability of the component are usually small 11. At high incident impact energies target perforation may occur and the passage of the impactor will generally result in petalling, cracking and spalling 12. Although such damage will degrade the load-bearing ability of the structure, its effects can generally be predicted using fracture mechanics principles t2. In composites, however, the ability to undergo plastic deformation is extremely limited with the result that energy is frequently absorbed in creating large areas of fracture with ensuing reductions in both strength and stiffness 1.13. Furthermore, the prediction of the postimpact load-bearing capability of a damaged composite structure is more difficult than for metals since the damage zone is generally complex in nature and consequently very difficult to characterize 12. The problem is further complicated by the lack of existing standards or established testing techniques for impact of composite materials. Much of the work published in the literature has been conducted on purpose-built machines using convenient specimen geometries. As a result, direct comparisons between different material systems is often very difficult and immediate conclusions are sometimes hard to draw. The objective of this article is to draw together the pertinent findings of many of the articles published in the field of impact of fibre-reinforced composites and thereby present a global view of the state-of-the-art. Initially, the paper will review briefly the techniques most frequently employed for assessing the impact 347

0010-4361/91/050347-16(~) 1991 Butterworth-Heinemann Ltd COMPOSITES. VOLUME 22. NUMBER 5. SEPTEMBER1991

response of composite materials. Following this, the effect of varying fundamental parameters such as fibre and matrix properties and stacking sequence on the dynamic response of these materials will be considered. Since a substantial amount of the work published in the literature relates to tests on carbon fibre composites, the main focus of this report will be on this material. However, work undertaken on other materials will be used in order to supplement and substantiate ideas and theories.

Specimen Notch ~I/-- ~pactor

Weig Inertiabar Straingauge [ --I

IMPACT TEST TECHNIOUESFOR COMPOSITE MATERIALS Ideally, the impact test fixture should be designed to simulate the loading conditions to which a composite component is subject in operational service and then reproduce the failure modes and mechanisms likely to occur. In simple terms, the impact problem can be divided into two separate conditions: low velocity impact by a large mass (dropped tool) and high velocity impact by a small mass (runway debris, small arms fire, etc.). The former is generally simulated using a falling weight or a swinging pendulum and the latter using a gas gun or some other ballistic launcher. However, as stated previously, due to a lack of experimental standards a wide variety of testing techniques is presently being employed in order to assess the dynamic response of reinforced plastics 1-4, 14-17, making direct comparison difficult.
In this section the more commonly used techniques will be presented and discussed as well as the problems associated with the ensuing data analysis.

Specimen [ __ Notch

Specimen or /I Straingauges

o ar e'P"t
b e

Low velocity impact Machines currently used for simulating the low velocity impact response of composite materials include the Charpy and Izod pendulums, the falling weight fixtures such as the Gardner and drop dart tests as well as hydraulic machines designed to perform both in-plane and out-of-plane testing at velocities up to 10 m s-1. 1. Charpy pendulum
Many of the early impact studies on composite materials were undertaken using the Charpy test method originally developed for testing metals 11'1s'19. The reason for this choice was the fact that the Charpy pendulum is both simple to use and can be instrumented, and therefore, in principle, can yield information on the processes of energy absorption and dissipation in composites. The test specimen is generally a thick beam, sometimes incorporating a notch at its mid point as shown in Fig. l(a). The specimen is supported in a horizontal plane and impacted by the swinging pendulum directly opposite the notch. The energy dissipated during impact is usually recorded by a dial on the test apparatus. Further information can be obtained by instrumenting the impactor with a strain gauge, thereby enabling the determination of the variation of the impact force with time. The energy absorbed during impact can also be determined by integrating the force/time signal.
348 COMPOSITES. SEPTEMBER 1991
Fig. 1 (a) Charpy impact test; (b) Izod impact test; (c) tensile Hopkinson-bar test; (d) gas gun

The test set-up suffers a number of disadvantages such as the fact that the load/time curves often contain high frequency harmonic oscillations resulting from the natural response of the impactor, etc. These effects can generally be filtered once the harmonic frequencies of the various components have been determined 2. As stated previously, the test specimen is a short, thick beam and is not therefore typical of engineering components. Further, the test is destructive, inducing failure modes that are not necessarily observed under low velocity impact loading on operational structures. Bader and Ellis.21 showed that the Charpy energy of carbon fibre-reinforced plastic (CFRP) varied with specimen geometry and doubt, therefore, the applicability of the technique. The Charpy test is only suitable for ranking the impact performance of continuous fibre

composites and as a first step in determining the dynamic toughness of these materials.

to geometries such as the DCB since the specimen may not be deforming in the same mode as a similar statically-loaded specimen24.

2. Izod test
The Izod impact test is shown schematically in Fig. l(b). The test set-up and procedure are similar to those outlined above. In the Izod test specimen is clamped in the vertical plane as a cantilever beam and impacted by a swinging pendulum at the unsupported end. The test suffers similar problems to those reported above and again is best suited as a tool for ranking the impact resistance of composite materials.

Intermediate and high rate impact 1. Hopkinson-bar technique


The Hopkinson-bar technique is similar to the previous test procedure in that it permits the determination of the variation of basic material properties as a function of strain rate. Several different types of Hopkinson-bar are currently employed; these include the punchloaded Hopkinson-bar, the compression bar, the tensile bar and the Hopkinson-bar shear test z~-27. The set-up and experimental procedures associated with these tests has been discussed in detail elsewhere25. Here, a brief description of the tensile test will be given. The test set-up for undertaking dynamic tensile tests is shown in Fig. l(c) ~. Here, specimens waisted through the thickness are bonded into slots in the inertia bar and input bar as shown in the diagram. This configuration is then placed within the weighbar. Strain rates approaching 1000 s-1 can then be achieved using gas-driven projectiles to accelerate the weighbar and, in turn, load the input bar 25. The inertia of the inertia bar then enables the test coupon to be loaded dynamically. Strain gauges bonded to the input and inertia bars enable the incident and reflected stress waves to be analysed and permit the determination of a dynamic stress/strain curve for the material. The Hopkinson-bar tensile test is destructive. Care has to be taken in order to ensure that the interface between the specimen and loading bars is good, otherwise a shear failure within the gripping section is likely to occur. Further, in order to minimize stress concentrations associated with the gripping area, relatively long specimens (approximately 60 mm) are required when testing composite materials .
28

3. Drop-weight impact tests


Here a weight is allowed to fall from a pre-determined height to strike the test specimen or plate supported in the horizontal plane. In general, the impact event does not cause complete destruction of the test specimen hut rebounds, enabling a residual energy to be determined if necessary. The incident velocity of the impactor can be determined from the equations of motion or by using optical sensors located just above the target. Frequently, the impactor is instrumented, enabling the force/time characteristics to be determined, and may also contain a displacement transducer in order to permit the determination of energy dissipation during the impact event. One of the advantages of this test with respect to the Charpy and Izod tests is that a wider range of (est geometries can be tested, thereby enabling more complex components to be tested. Although testing is generally undertaken using a hemi-spherical impactor, it is possible to use other impactor shapes such as blunt cylinders or sharp points.

Variations on the drop-weight theme include the Gardner test where a hemi-spherical impactor strikes a small diameter circular plate and the driven dart test where a hemi-spherical probe is driven into the specimen at a pre-determined rate 22.

2. 4. Hydraulic test machines


In recent years a number of workers have used hydraulic test machines for assessing the deformation and failure characteristics of materials at high rates of strain 23'24. Here, test geometries such as tensile dogbone specimens or double cantilever beam (DCB)type specimens can be tested over a wide range of strain rates The strain history of the specimen can be measured using bonded strain gauges or an optical transducer. If a strain gauge or any other displacement measuring device is bonded to the specimen, the strain rate sensitivity of the adhesive should be considered. The advantage of this technique is that the test specimens permit the evaluation of basic material properties such as tensile strength, modulus and interlaminar fracture toughness without the contact effects associated with falling weight impact. Extreme care has to be taken in order to ensure that the mass of the load cell and gripping system are as low as possible since inertial effects resulting from these components may conceal the true material response24. Caution should also be exercised when applying fracture mechanics principles

Gas gun impact testing

Impact testing at ballistic rates of strain can be achieved using a high pressure gas gun such as that shown in Fig. l(d) ~9. Typically, a gas such as nitrogen is fed to a chamber located at one end of the barrel. Here the gas is restrained by a plastic diaphragm. When the gas has reached a pre-determined value the diaphragm is burst (by electrical heating or a mechanical puncturing device), accelerating a projectile down the barrel to strike a specimen or component supported vertically. The velocity of the impactor can be determined just prior to impact using optical sensors3 or by using a simple break-wire technique 29. Generally, the test is not completely destructive but frequently results in large-scale damage and/or target perforation. Until recently the technique suffered the disadvantage that very little information could be obtained from the test itself. However, instrumented gas guns have now been developed, enabling force/displacement histories to be measured and the impact event to be analysed in more detail31,32. Gas guns can be used to test large structures and are therefore useful for assessing the high velocity impact response of composite materials.

COMPOSITES. SEPTEMBER 1991

349

Conclusions At present, no acceptable standard testing procedures are available for impact testing of composite materials. Consequently, a wide variety of testing procedures, specimen geometries and data reduction techniques are presently being employed. Pendulum techniques such as the Charpy and Izod tests often require specimen geometries that are not representative of component dimensions and so are essentially suitable only for ranking the impact response of composites. Dropweight rigs and gas guns offer more representative approaches for assessing the impact response of these materials. Greater use of instrumented impactors has led to a deeper understanding of the processes of energy absorption and dissipation in these materials. More recently, hydraulic machines have been used to examine the response of composite materials at high rates of strain. Although somewhat more expensive than other, more conventional techniques, such machines are ideally suited for analysing the rate dependency of basic material properties. INFLUENCE OF CONSTITUENT PROPERTIES ON THE IMPACT RESPONSE OF COMPOSITE MATERIALS In sirqple terms, a fibre-reinforced composite is composed of three constituents: the fibres, the matrix and a fine interphase region (sometimes referred to as the interface) responsible for assuring the bond between the matrix and fibre. The manner in which the material deforms and fractures depends upon both the chemical and mechanical properties of these three constituents. Detailed analyses using both optical and scanning electron microscopes have identified a number of failure mechanisms in fibre-reinforced composite materials29'3336. These include delamination, intralaminar matrix cracking, longitudinal matrix splitting, fibre/matrix debonding, fibre pull-out and fibre fracture. The relative energy absorbing capability of these fracture modes depends upon the basic properties of the constituents as well as the loading mode. Typical values measured for a number of continuous

fibre composites are given in Table 1. Generally speaking, failure modes that involve fracture of the matrix or interphase region result in low fracture energies whereas failures involving fibre fracture result in significantly greater energy dissipation. The following sections will investigate the role of the material constituents independently and examine the influence of varying their properties on the overall impact resistance of the composite.

Fibre The role of the fibres in a composite structure is extremely important since they are responsible for bearing a significant percentage of the applied load. At present, many types of fibre are available. In aeronautical applications these include carbon, glass, and KevlaP~ fibres. Within each of these categories fibres exhibiting a wide range of mechanical properties are available. Unfortunately, it is often difficult to separate the effects of mechanical properties (such as strength and stiffness) from those arising from geometrical factors (such as fibre shape and diameter) and interfacial properties (such as the strength of the chemical bond between fibre and matrix). Early studies36, in which the relative performance of a number of continuous fibre composites was examined, suggested that the Cliarpy impact resistance of S-glass and Kevlar~ fibre-reinforced composites was over five times greater than that of a Modmor II carbon fibrereinforced composite. Beaumont et al. is examined the Charpy load/time traces of a number of materials and showed that the curve corresponding to an HMS carbon composite was extremely brittle, failing catastrophically at maximum load. The E-glass and Kevlar~ 49 composites failed in a more progressive manner indicative of energy dissipation through delamination, splitting and other failure processes. The authors quantified the Charpy load/time traces by defining a ductility index (DI), this being the ratio of the energies associated with the crack propagation phase (the area after maximum load) and the initiation phase (the area

Table 1. Typical values of the energy absorbing capability of various continuous fibre composites for different failure modes Failure mode Material Typical fracture energy (kJ m-2)
0.1-1 3.8 0.1 2.2 20 60 128 26 800 6

Reference

Splitting Delamination
Transverse fibre fractu re Fibre pull-out Debonding

Type II CF/epoxy AS4/PEEK T300/epoxy IM6/PEEK

11 60 102 62 72 72 60 104 105 106

Treated CF/epoxy Untreated CF/epoxy AS4/PEEK


CFIpolyester CF/bismaleimide CF/epoxy

350 COMPOSITES. SEPTEMBER 1991

up to maximum load). The resulting ductility indices for the Kevlar-49, E-glass and HMS carbon/epoxy composites were 23, 0.4 and 0.0 respectively, indicating clearly the superior energy absorbing capability of the Kevlar fibre. Hancox 37 and Bader and Ellis 21 compared the low velocity impact response of composites containing type I and II carbon fibres. In both cases it was shown that the materials containing type II fibres (higher failure strength) offered superior impact resistance. Similar results have been observed following low velocity impact tests on E- and S-glass fibre composites 15. In an attempt to better understand the fundamental parameters controlling the processes of energy absorption and dissipation in composite materials, Chamis et al. 3s undertook Izod impact tests on a wide range of systems. They concluded that flexure and interlaminar shear deformations are dominant energy-absorbing mechanisms in composites and that the area under the material's linear stress/strain diagram represents a useful approach for predicting the impact resistance of a composite. Essentially, composites with large areas under the stress/strain curve are more effective energy absorbers. Fig. 2 presents the experimental data obtained by Chamis and Sinclair39 plotted as a function of the energy absorbing capability of the fibre as determined by the energy under the static tensile stress/strain curve. An examination of the data suggests a possible relationship between these two parameters, with materials containing fibres with a greater strain energy absorbing capacity offering improved lzod energies. It appears, therefore, that this technique forms a useful guide for assessing and evaluating the impact resistance of composite materials. This technique also explains the results of Hancox 37 and Bader and Ellis 2~, reported above, who found that composites based on type II carbon fibres outperformed type I carbon fibre composites. The former has an elastic energy absorbing capability some four times greater than the latter. Although the above approach is valuable for evaluating the impact resistance of a composite, a complete analysis should take into consideration energy dissipation in failure processes such as fibre/matrix debonding
40,

and fibre pull-out. Beaumont 4 gives expressions for work associated with micro-mechanical fracture processes such as debonding and fibre pull-out. The wot'k for debonding is given as: Wd = ad2o2fld24Ef (1) where d = fibre diameter, of = failure strength of the fibre, ld = length of the debonded zone and Ef = fibre modulus. The work to pull-out is given as:
Wp = ~d~Fg2 4

(2)

where lc = critical transfer length (-- otd/2x) and x = constant frictional shear stress. Beaumont concluded that post-debond fibre sliding is the primary energy absorbing mechanism in glass fibre composites, whereas fibre pull-out is responsible for much of the toughness in a carbon fibre composite. An examination of Equation (2) indicates a strong dependency of work to pull-out on fibre diameter. In theory, increasing the diameter of the reinforcing fibre should result in a composite with an improved resistance to pull-out and perhaps improved toughness 1L34. Morris and Smith41 reported results that appear to support a fibre diameter dependence in continuous fibrereinforced composites. Their conclusions are based, however, on differences between glass and boron fibre composites with different fibre volume fractions and are not, therefore, conclusive. In recent years, fibre manufacturers have been improving the strain to failure of carbon fibres by reducing their diameter. Typically, the first generation of carbon fibres such as T300 and AS4 had diameters of 7--8 ~tm. More recent fibres such as IM6 have diameters of approximately 5 tim. By improving the strain to failure of the fibres in this way the manufacturers have also improved the strain energy absorbing ability of composites and thereby improved their impact resistance. Davies et al. 42 showed that the interlaminar fracture toughness of IMr/polyetheretherketone (PEEK) (superior strain energy absorbing capability) is superior to that AS4/PEEK. Curson et al. 4 drew similar conclu#ons following low velocity impact tests on these materials. Under conditions of low velocity impact loading, it appears, therefore, that the improved toughness gained by increasing the fibre diameter is offset by the reduction in fibre failure strain, resulting in a loss of strain energy absorbing capability. It may be, however, that increasing the fibre diameter might lead to improvements in the high velocity impact response of composite materials. Under such loading conditions, the target response is more localized with energy being dissipated over an area immediate to the point of impact. Consequently, the overall energy absorbing ability of the structure is less important and local mechanisms such as fibre pull-out and fracture become dominant. As well as influencing the impact resistance of a composite, the properties of the fibre also have a significant effect upon the residual load-bearing capability of the material. Composite systems that offer excellent impact resistance do not necessarily exhibit superior postimpact residual properties. For example, polyethylene

35
I

30 ~25 ~E
2O

.E

Q.

15-10

0
0.0

I t{,
0.2

I,

,ii

,,,

1,

0.4

0.6

0.8

1.0

1.2

Normalized strain energy Fig. 2 Variation of Charpy impact energy with normalized strain energy absorbing capacity of fibres (data taken from Ref 39)

COMPOSITES. SEPTEMBER 1991

351

fibre composites are capable of absorbing large amounts of energy before failure but still offer relatively poor compressive properties due to the low stiffness of the fibres themselves. In most circumstances a compromise is required. The following section considers briefly the effect of fibre properties on the post-impact strength of composite materials. 1. Effect of fibre properties on post-impact residual strength In recent years a concerted effort has been made by the manufacturers of high performance fibres to improve both the short- and long-term mechanical properties of advanced composites. Much of this improvement has been achieved by increasing the strain to failure of the reinforcing fibres. In the previous section, the energy absorbing capacity of the fibres was identified as an important parameter in determining the level of damage incurred in a composite laminate. Often, materials that satisfy this condition also offer excellent residual properties. Cantwell et al. 43 showed that this was the case for low and high strain carbon fibre composites. An AS4 carbon fibre composite with a superior strain energy absorbing capacity than that of an XAS carbon fibre composite offered superior residual properties, Fig. 3. This is not always the case, however. For example, if the stiffness of the fibre is very low and its strain to failure high, a composite containing these fibres will be capable of absorbing large amounts of energy but will exhibit poor residual compressive properties. In order to overcome this, hybrid composites are frequently used, combining the energy absorbing capability of low modulus fibres with stiffer fibres capable of resisting compressive loads**. As stated previously, many of the latest generation of composites are based on fibres with smaller diameters. Since the compressive strength of a composite depends upon the stability of the fibres, it would be expected that smaller diameter fibres would result in a material with poorer compressive properties. This indeed appears to be the case 45. However, reducing the fibre diameter increases the energy absorbing capability of the composite, resulting in lower levels of damage for a
1200 ~- 1000

given incident energy. The reductions in plain compressive properties of the composite appear to be offset by the reduction in damage area. 2, Conclusions

The properties of the fibres in a continuous fibre composite have a significant effect on the impact resistance and subsequent load-bearing capacity of components made from such materials. For low velocity impact loading, the ability of the fibres to store energy elastically appears to be the fundamental parameter in determining impact resistance. Kevlar fibres, which have large areas under their stress/strain curves, offer excellent impact resistance. The role of the fibre diameter is not completely clear. A simple pull-out model suggests that composites with larger diameter fibres should be inherently tougher. However, current trends are towards smaller diameter fibres offering higher strains to failure. Any reduction in toughness is thereby hidden by the increased energy absorbing capacity of the fibres. Matrix The polymeric matrix in a fibre-reinforced composite serves to protect, align and stabilize the fibres as well as assure stress transfer from one fibre to another. In general, both the stiffness and strength of the matrix are considerably below those of the reinforcing fibre. The latter is therefore responsible for carrying most of the applied load in a composite component. The role of the matrix is nevertheless critical. For example, damage to the matrix such as impact-induced delamination can reduce the load-bearing capabifity of the composite by up to 50% 46. As a result of this relatively poor behaviour, much work has been undertaken in recent years in an attempt to identify the fundamental matrix properties that influence the impact resistance of composite materials. Since the first generation of matrix systems for advanced composites lacked toughness, a number of techniques have been developed to impro~,e the toughness of these materials. These include: the use of plasticizing modifiers11; the addition of rubber particles such as carboxylterminated butadiene-acrylonitrile (CTBN)47--49.; the addition of thermoplastic particles such as polyethersulphone (PES)and polyetherimide

?
~ 600
~, 400
n.-

e-

~.~ ~ " ' - o

.....

o..

Tension

(pE0~52;

a reduction in the cross-linking density of thermosets such as epoxy resins53'54; the use of thermoplastic matrices such as PEEK; and
Compression/I'I"
I

200

the inclusion of thin, tough layers at ply interfaces 5,5s-57.


I
I

10

Impact energy (J) Fig. 3 Variation of tensile and compressive residual strength w i t h impact e n e r g y for l o w and high strain fibres (Ref 43): e, high strain fibres; o, l o w strain fibres

Early work by Bradshaw et al. 11 showed that adding a plastidzer to Epikote 828 epoxy resin increased the Mode I fracture toughness by over two orders of magnitude. When used as a matrix system in a carbon fibre composite, increases in toughness did result; however, in this case the Izod impact energy was

352

COMPOSITES. SEPTEMBER 1991

improved by only 25%. This disappointing transfer of toughness was explained by the fact that the Izod test induces crack propagation across fibres rather than between them. Improvements in the interlaminar toughness and resistance to splitting are hidden, therefore, by the higher energy fracture energy associated with transverse fibre failure. Husman et al. 17conducted high velocity impact tests on a limited number of Modmor II carbon fibre composites. Their data suggested that the impact resistance of these materials did not depend upon the properties of the polymeric matrix. In a more detailed analysis by Williams and Rhodes 49, the impact resistance of 24 modified and unmodified carbon fibre/epoxy composites was examined. Their experimental analysis showed that both the level of damage incurred as well as the residual compressive properties of the laminates varied enormously. It was found that the brittle laminates tended to fail by extensive delamination whereas the tougher systems failed in transverse shear near the impact location. The authors concluded that the tensile performance of the neat matrix has a significant influence on the impact behaviour of a composite structure. For improved impact resistance, the strength of the matrix should exceed 69 MPa and its strain to failure should be greater than 4%. Finally, in order to ensure adequate compressive strength, the shear modulus should be greater than 3.1 GPa. Hirschbuehlers7 examined a large number of plain and interleaved composite systems under a variety of different loading conditions in an attempt to relate matrix properties and post-impact compressive strength. It was found that the residual compressive properties of the composites increased with the flexural strain to failure of the pure resin, Fig. 4. Hunston47 analysed data from three sources in an attempt to identify a link between matrix properties and composite fracture toughness. He identified a definite correlation between the resin Mode I fracture toughness and composite interlaminar fracture energy as measured by the DCB specimen, Fig. 5. With brittle

polymers, the resin toughness is fully transferred to the composite, whereas with tougher polymers the resin toughness is only partly transferred to the composite. In the latter it is proposed that the presence of the fibres restricts the crack-tip plastic zone size, thereby reducing the positive effect of the tougher matrices. Masters3s'~ extended this approach by conducting Mode I, Mode II and compression after impact (CAI) tests on a number of epoxy and bismaleimide-based carbon fibre composites. He showed that no correlation existed between the Mode I interlaminar toughness and the CAI properties, Fig. 6, whereas a very good agreement was found between the Mode II resistance and residual compressive strength, Fig. 7. Similar observa tions have been made by other workers58 . It is clear that the matrix in a flexurally loaded composite will be subjected to a large Mode II component and that the forward shear properties of the matrix will be important in determining the level of damage incurred. It is somewhat less clear why the residual compressive properties of the composite should be Mode II controlled since the failure process is undoubtedly complex containing a significant Mode I component. Master's results are impressive nevertheless. Materials that satisfy the above condition and therefore
I

3
o

E -'-

J
0 O0

/ o

-~ Thermoset o Experimental
0

O J ~

'~ E
O

Toughened thermoset + Thermoplastic


I L

Resin fracture energy (kJ m -2 ) Fig. 5 Variation of composite Mode I interlaminar fracture energy with resin fracture energy (Ref 47)
500 ~-

a. ,'-

350 I 300 250 00

.."

E 450 7
400 ~ AS4/907 AS4/1808-fi(m E C6000/1827-film A [M6/1808 AS4/1808 AS4/1808-fitm C C60O0/1827 ] M 6 / 1 8 0 8 - f i l m A AS4/1808-filrn A

~ == 350

_= 300
250

2I
150
'~ 50 0 0.0 ,

~ l I
$ ~:

150 100 50 0 0
I 1 AS4/3502

A$4/3502-fi{m A

"o

, I , , , I I ~ , I. ,. , , I , , , I , L , 2.0 4.0 6.0 8.0 10.0 12.0 Neat resin flexural strain to failure (%)

100

200

300

400

500

Fig. 4 Variation of residual compressive strength of impactdamaged composites with neat resin flexural failure strain (Ref 57)

Residual compression strength after impact (MPa) Fig. 6 Variation of residual compression strength after impact with Mode I strain energy release rate (Ref 56)

COMPOSITES. SEPTEMBER 1991

353

4000
AS4/1808:film E

E 3500

~ 3ooo
_~ 2500
2000 c 1500 1000

IM6/1808-film A AS4/3502.film A

eAS4/1808-fiim A

AS4/1808-film C IM6/3100-film E IM6(3100-21-film E C6000/18274ilm E AS4/907

tive to compressive loading since impact-generated delaminations tend to reduce the stability of the load-beating plies resulting in premature failure through buckling. In the previous section it was shown that composites with high Mode II interlaminar fracture toughnesses offer superior CAI propertiess5,56. It appears that the majority if not all of this improvement in CAI behaviour results from the lower levels of damage incurred during impact. The residual tensile properties of toughened composites do not appear to be significantly better than those of standard epoxy systems, Fig. 8. This results from the fact that composites with tougher matrices tend to be more notch-sensitive due to reduced splitting and delamination around stress concentrations such as notches or damage6s. Toughening composites using elastomeric particles reduces the level of delamination and therefore enhances residual compressive properties, Fig. 9. However, the presence of such inclusions often reduces the glass transition temperature of the matrix material which in turn reduces the hot-wet properties of the composite. In many situations a compromise is therefore necessary.
1400

A54/1808 ....... AS4/1808 l M o / I ~ u ~ C6000/1827 AS4/3502 I M 6 / 3 1 0 0 e I M 6 / 3 1 0 0 - 2 1

0 0

t 100

I 200

i 300

I 400

I 500

Resin compression strength after impact (MPa) Fig. 7 Variation of residual compression strength after impact w i t h M o d e II strain e n e r g y release rate (Ref 56)

offer superior impact properties include thermoplasticbased composites and interleaved laminates. In recent years considerable interest has been generated by carbon fibre-reinforced PEEK (APC2), a semicrystalline thermoplastic compositea'~'59-62. Interlaminar fracture testing and impact loading have shown that this material offers excellent static and dynamic toughness and is capable of absorbing a considerable amount of energy whilst incurring only small amounts of damage4,63,64. Scanning electron micrographs of the fracture surfaces indicate extensive drawing and plastic flow65. Another advantage of this material is that its thermoplastic matrix allows rapid repair using fusion techniques such as the hot press technique66. Here, impact damage can be reduced or removed by simply heating the component to a temperature above the melting point of the matrix, reforming and cooling. The high velocity impact response of carbon fibre/PEEK has received very little attention. Initial testing has suggested that its high velocity impact response is perhaps relatively poor. Dan-Jumbo et al. 59 showed that beyond a certain velocity threshold, APC2 experienced a sudden drop in flexural strength. Similar observations have been observed by Morton and Godwin63 following ballistic impact tests on this material. This will be discussed in more detail in the section on rate effects. These observations suggest, therefore, that care should be exercised when attempting to relate static properties such as interlaminar toughness and strength to characterize tlynamic properties such as impact resistance.
Polymer interleafing involves the use of high toughness films or layers at ply interfaces in relatively brittle materials. The inclusion of such layers increases the laminate's interlaminar fracture toughness56 as well as reducing the level of damage incurred for a given incident energy67. The load-bearing properties of damaged intedeafed composites are significantly superior to those of conventional epoxy composites67. Interlayer technology is still in its infancy; however, early results are very favourable and the technique offers enormous potential. 1. Effect of matrix properties on post-impact residual strength

<.
1200' 1000
80o ~ 600

O - -- -- Tough epoxy Brittle epoxy

"0 .....

g
n-

4o0
200

0.0

2.0

4.0

6.0

8.0

,IL J ,I, f ~ 10.0 12.0 14.0

Impact energy (J) Fig. 8 Variation of residual tensile strength with impact energy

for tough and brittle epoxy-based composites (Ref 75)

1000
~
- -

900,

e-

800, 700 600

Toughened epoxy o Brittle epoxy

~
~

s~
400

72 8

200 100 0 a 2 4I f ~ 6 8 Impact energy (J) 110 112 14

Impact-damaged composites are probably most sensi354 COMPOSITES. SEPTEMBER 1991

Fig. 9 Residual compressive strength of toughened and brittle epoxy-based composites as a function of impact e n e r g y (Ref 75)

2.

Conclusions
(3.

1000 ~ 800 " --600

It is clear that matrix properties play a significant role in determining the impact resistance and subsequent load-beating capability of a fibre-reinforced plastic. At present, a significant effort is being made to improve the post-impact compressive properties of composites. It appears that materials with high Mode II interlaminar fracture toughnesses offer superior compression strengths after impact. Care should be exercised, however, when correlating data obtained under static loading conditions with results from dynamic tests.

Surface-treated fibres Penetration

0 Untreated fibres

-=E

'\l
/
0 2 4

I/J

..- - O -

Interphase
The strength of the bond between the matrix resin and the fibre reinforcement is a controlling factor in determining the mechanical performance of most polymer composites. In general, the surface of the fibres is treated by an oxidative process in order to improve the level of adhesion between matrix and fibre. Initially, this interracial zone was considered as being a twodimensional surface with effectively zero thickness. However, more recent studies have shown that this region is in fact three-dimensional, having its own distinct properties 69. Studies have shown that varying the level of surface treatment applied to a carbon fibre composite can change the mode of failure as well as many fundamental mechanical properties H'69-7z. Composites with low levels of fibre surface treatment fail at relatively low stresses when loaded transversely to the fibres, leaving smooth fibres on the fracture surface. Increasing the level of treatment applied to the fibres increases the transverse failure stress and failure occurs within the matrix, i.e., the interphase region is no longer the weakest link in the composite. Rogers e t al. 71 showed that improving the fibre/matrix bond strength in a carbon fibre-reinforced epoxy resulted in a fourfold increase in the incident impact energy required to initiate damage. At higher energies the load-bearing properties of composites with surfacetreated fibres drops dramatically until the perforation limit is reached72. Bless and Hartman 73 showed that the perforation threshold energy in a surface-treated composite is significantly lower than that of a similar untreated laminate. This behaviour has been explained by Dorey~'2, who showed that the transverse fracture energy of a composite, a fundamental parameter for determining resistance to penetration and perforation, depends strongly upon the fibre/matrix bond strength. Carbon fibre-reinforced epoxies with untreated fibres offer transverse fracture energies as high as 60 kJ m-2 (Ref 72). Transverse failure in composites with high levels of fibre surface treatment absorbs considerably less energy, with quoted ~2 transverse fracture energies being as low as 20 kJ m-2. At energies levels above that required to achieve perforation, damage in a surfacetreated composite tends to be localized around the point of impact, often taking the form of a clean hole 71. The post-perforation residual properties of treated composites are generally superior to those of untreated composites as shown in Fig. 10.

10

12

14

Incident energy (J) Fig. 10 Residual flexural strength vs. impact energy for ballistically impacted surface-treated and untreated carbon fibre composites (Ref 72)

McGarry et al. TM attempted to enhance the impact resistance of T300/MY-720 carbon fibre/epoxy by coating the surfaces of the fibres with a thin layer of CTBN rubber. Gardner impact tests showed that treating the fibres in this way improved the threshold energy for first damage significantly. Further, above this threshold the level of damage for a given impact energy was less in the modified composite.

1. Effect of interphase properties on post-impact residual strength


Dorey 75 has shown that increasing the strength of the fibre/matrix bond increases the interlaminar shear strength (]LSS) of the composite until a plateau is reached beyond which point no significant increase is possible, Fig. 11. Over this range of treatment levels the notched tensile strength falls dramatically. Increasing the ILSS in this way suppresses the formation of delaminated zones in the region of stress concentrations, rendering the material more notch-sensitive. Consequently, even though surface treatment of the fibres reduces the level of damage for a given energy, the increased notch-sensitivity of the laminate results in poorer residual tensile properties, Fig. 1 2 7 5 . Conversely, treating the fibres improves the post-impact compressive properties as shown in the lower part of Fig. 12. Clearly, the level of surface treatment applied to the fibres in a multidirectional composite will depend upon the operational conditions the component will encounter. In general, a compromise is sought in which the fibres are given intermediate levels of treatment.

2.

Conclusions

The level of treatment applied to the surface of the fibres in a composite material has a significant effect upon both its impact resistance as well as its residual load-carrying capability. In general, impact on composites with low levels of fibre surface treatment generates large areas of splitting and delamination with severe effects on the compressive properties of the material. Localized impact loading on highly treated fibre composites results in a smaller, more localized damage zone, a lower perforation threshold and improved

COMPOSITES. SEPTEMBER 1991 355

'[
T/T100

Fibre stacking sequence


Composite materials offer a unique advantage in that properties such as strength and stiffness can be tailored to meet specific design requirements through a careful selection of the fibre stacking sequence. Considerable work has shown that the impact resistance of composite materials also depends upon the specific order in which the plies are stacked6'~'76-s. For example, unidirectional composites having all their fibres aligned in one direction fail by splitting at very low energies and are therefore highly unsuitable for applications where impact loading might occur 11. Following impact tests on a series of (0, +/-45 ) laminates, Dorey 1~as well as Morton and Godwin 63 showed that composites having + / - 4 5 surface plies offered a superior impact resistance and improved residual strengths. It was suggestedsl that + / - 4 5 plies increased the flexibility of the composite, thereby improving its ability to absorb energy elastically. Further, placing such plies on the surface of a composite serves to protect the loadbearing 0 plies against damage induced by the imping, ing projectilesl. These ideas were supported by Stevanovic et al. s, who conducted instrumented Charpy tests on a series of multidirectional T300 carbon fibre composites. They showed that (+/-45 ) composites were capable of absorbing considerably more energy than (0,90), (0,+/--45 ) and (00,900,+/-45 ) laminates. A somewhat different approach was adopted by Nolet and Sandusky in the design of a composite leading edge for the A10 Thunderbolt 8z. They proposed that the leading edge be very stiff in order to 'split off' the impinging bird with little or no loss in strength or stiffness. They suggested that in order to achieve this requirement, a high percentage of carbon fibres should be oriented in the spanwise direction of the aircraft's leading edge. The authors s3 showed that damage initiation in a series of (+/-45 ) laminates subjected to low velocity impact depended upon the thickness and therefore the stiffness of the composite. Initial failure in thin, flexible targets occurs in the lowermost ply as a result of the tensile component of the flexural stress field. Damage in thicker, stiffer targets initiates at the top surface due to
1.6 '1,4
.....
~ ~ . "~_=t .- - . _--

1.0

On/OnlO0

0.5

0 0

I 100

I 200

I 700

Surface treatment (%)

Fig. 11 Variation of interlaminar shear strength and notched strength of carbon fibre composites as a function of fibre surface treatment (Ref 75)

1.0 0~
A tO n

Tension

Treatment level 0% o . . . . . 5% " - - - - 100%

t3 v .c 0.6 ~,+\ \~,\ -~ 0.4 '~,,. ~.


"0
X~O~ ~'II~ ~+

............ ".. . . . . . . .. .. .. .
I

o ... +
I

Lower [ Upper surf~rface

1,2
1.0
O3

0.2 0.0

/
Compression
I I

o.8
0.6 0.4

10 E

Impact energy (J)

Fig. 12 Residual strength of treated and untreated carbon fibre composites vs. impact energy (Ref 75)

compressive properties. However, the increased notch sensitivity associated with fibre surface treatment results in a reduction in the post-impact tensile strength of the material. The level of treatment applied to the surface of the fibres will depend, therefore, upon the desired application.
356 COMPOSITES. S E P T E M B E R 1991

0,2 0.0
0
,,l..,n , , , l . . , l , , , J , , , l . , , l , , , f . , , l , , ,

3 4 5 6 7 Target thickness (mm)

9 10

Fig. 13 Low velocity impact energy to initiate damage vs. target thickness for (+/-45 ) CFRP composites (Ref 29)

the contact stress field. This is shown in Fig. 13. This curve clarifies Dorey's sl claim that increasing the flexural stiffness of a target, for example, by placing fibres on the surface of a laminate, can enhance its impact resistance. This is true for the range of stiffnesses where initial failure occurs at the top surface of the component. In more flexible targets, however, reducing the flexural stiffness may precipitate failure at a lower incident energy. A detailed study by Hong and Liu 6 identified fundamental aspects in the development of damage in glass fibre-reinforced plastic (GFRP) subjected to high velocity impact loading. They showed that increasing the angle q in a (~,q,0~) laminate resulted in greater delamination-type damage for a given incident energy, Fig. 14. Increasing q in this way also had the effect of reducing the first damage threshold energy. The authors als0 showed that for a given energy, increasing the thickness of the GFRP target resulted in an increase in delaminated area. This increase in damage area may result from the reduction in the target's energy absorbing capability as proposed by DoreySL Lius4 extended this work by developing a simple model for predicting the likely delamination sites in a number of different composites. It was suggested that delamination in multi-angle composites is more likely to occur at interfaces where the mismatch in bending stiffness is greatest, for example, between +/--45 plies. Liu showed experimentally that the level of delamination in a glass/epoxy composite increased as angle q in a (0,q) laminate increased, i.e., as the bending stiffness mismatch increased This evidence suggests that if delamination needs to be suppressed, laminates with sudden large changes in fibre direction should be avoided. Other techniques to reduce impact-induced delamination include the use of woven fabrics46"85, hybridization 16'a6-88 (for example, carbon fibres with Kevlar fibres) and three-dimensional stitchings9'9. The first of these techniques involves replacing the unidirectional +/-45 plies in a multidirectional composite by a +/-45 woven fabric. The three-dimensional nature of the fabric helps suppress the formation of delaminated zones at this critical interface.

The impact resistance of carbon fibre composites can be enhanced considerably by incorporating plies of lower modulus fibres 16,s6-s8. In order to assure compatibility, the matrix resin is usually the same in the two or more constituent materials. Hancox and Wellsa7 showed that the Izod impact energy of an HT-S carbon fibre composite could be increased by 500% through hybridization with E-glass fibres. As well as reducing the basic price of the composite, the addition of the glass fibres was found to change the mode of fracture from a clean break to a delamination-type noncatastrophic failure Similar conclusions were drawn by Helfinstine 16following Charpy impact tests on KevlarT300 carbon fibre hybrids. However, the absolute magnitude of the increases was less impressive than that reported by the previous workers for GFRP. Dorey e t al. ~ assessed the high velocity impact response of a number carbon-Kevlar hybrid laminates. They showed that the addition of the lower modulus Kevlar fibres increased the threshold energy for the onset of damage by up to four times. At present many workers are assessing the feasibility of using weaving and braiding techniques in order to improve the damage tolerance of fibre-reinforced polymer composites s9-93. Su 93 conducted Mode I delamination tests on both stitched and non-stitched AS4 carbon fibre/J1 (a semi-crystalline thermoplastic). His results showed that stitching with Kevlar fibre resulted in a 100% increase in interlaminar fracture toughness. Instrumented drop-weight impact tests on a number of 2- and 3-D composites indicated that the latter offered a superior impact resistance, the presence of the third dimension reinforcement served to inhibit the propagation of delaminated zones. Fabricating 3-D structures is clearly more expensive and time consuming than constructing with conventional 2-D prepregs. In order to reduce these costs, it has been proposed that a selective procedure be adopted, that is, 3-D reinforcement be used at critical ply interfaces or at component edges 93.

1. Effect of fibre stacking sequence on post-impact residua/ strength


In this section the role that the fibre stacking sequence plays in determining the residual properties of impactdamaged composites is discussed As outlined above, much of the work published in the literature concerns the residual compressive properties of damaged composites since this is considered to be the most critical form of loading condition Certain conflicts may exist, however, when considering the optimum fibre stackin.8 sequence for residual compressive strength. DoreyTM suggested that for improved impact resistance the +/--45 fibres should be located on the outermost surface of the composite. This may not be an ideal stacking sequence for stability in compression Here, stiffer laminates, for example, those with surface 0 fibres, are better suited to in-plane compressive loading. Nevertheless, Morton and Godwin6 3 have shown that an APC2 ( ~ , +/-45)2s laminate offers inferior properties to those of a (+/-45,~,+/--45,0)s plate, Fig. 15.

A cq

4 I
35 30 25 20 15 10 0

o =
[3

[ 05/90s/05 [ 0s/605/05 [ 05/455]0s [ 05/30s/05 [ 05/15~/0~

] ] ] ] ] ++0~]
+(

oo

r~

lid

C3 5 10 15 20 25 30 Impact energy (J) 35 40 45 50

Fig. 14 D e l a m i n a t e d area vs. i m p a c t energy for i m p a c t e d (0s*,05",05") GFRP l a m i n a t e s (Ref 6)

COMPOSITES, SEPTEMBER 1991 357

1400

i 45 outside

1200 I / ~ 0 outside I000 r

==
800 "

L
i E o o 6OO

they split and fail at low energies. The mismatch in bending stiffness between two plies appears to have a significant effect upon the level of damage incurred at that interface. Damage appears to be greatest where ply orientation changes of 90 occur. This suggests that for containment of damage laminates with abrupt changes in fibre direction should be avoided. Other ways to suppress damage include the use of woven fabrics, the use of hybrid composites or stitching at desired locations.

/k
4O0

Jk

Geometry
Geometry is a fundamental parameter in determining the impact response of a composite componenta,rt A5,63,75,83.Low velocity impact tests on CFRP have shown that the mode of failure in a simple beam may vary depending upon its span-to-depth ratio. Short thick specimens tend to fail in an interlaminar mode whereas as long thin beams failed in a flexure 11. Broutman and Rotem 15 showed that increasing the size of a GFRP beam increased its energy absorbing capability under low velocity impact conditions. However, doubling the size of the beam did not result in an equivalent increase in energy absorption. Similar tests o n CFRP 83 have shown that both the low velocity first damage threshold energy and perforation limit of a CFRP (0,-I-/---45 ) beam increased linearly with increasing beam length. These results indicated that the target's ability to absorb energy in elastic deformations determines its low velocity impact resistance. However, increasing the volume of the target does not necessarily increase its impact resistance, for example, small beams may be capable of absorbing greater energy than a large circular plate s3. Further, the process of failure in a large component often differs from that in a simple laboratory specimen. Clearly, care needs to be taken when using small, simple specimens to characterize the impact response of larger, more complex structures.

r~ 200 i 0 0 5 10 15 Incident impact energy (J)

Fig. 15 Effect of placing 45 plies on the outer surface of a 16-ply carbon fibre/PEEK laminate (Ref 63!
1000

(3 Non-woven Mixed-woven

.~ 8o0 Qe-

g, coo
e--

.o_ 400
e~ E o ~ 200

A W

2 3 Impact energy (J}

45) CFRP composite with a woven fabric (Ref 46)

Fig. 16 Effect of replacing the + / - 4 5 plies in a 16-ply (0,+/-

In the previous section it was stated that the use of woven +/--45" fabrics in (0", +/--45 ) laminates serves to reduce the overall level of delamination under impact loading. The subsequent residual strengths of the mixed-woven composites were superior to those of the standard material manufactured from unidirectional plies, Fig. 16. Similar improvements in residual strength have been noted following impact on stitched carbon fibre composites 92. Compression after impact tests on a number of AS4/3501--6 laminates92 showed that stitched laminates offered residual strengths up to 100% greater than their unstitched counterparts. One of the disadvantages of this process is that the undamaged compressive strength of the material is reduced by up to 20% 92.

High velocity impact tests on CFRP indicated that the areal geometry of the target is less important at high rates of strain 93. Ultrasonic C-scans of impacted specimens showed that the level of damage in a small, 50 mm long beam was the same as that in a 150 mm coupon, Fig. 17. This suggests that high velocity impact loading by a light projectile induces a localized form of target response in which much of the incident energy of the projectile is dissipated over a small zone immediate to the point of impact. Tests on large plates have substantiated this claim 93 and it appears that under certain conditions small simple coupons can be used to characterize the high velocity/low mass impact response of composite structures. Few workers have undertaken tests on full-size engineering components primarily as a result of the high costs involved. Gause et al. 94 showed that the curvature of the test component or structure influences the level of damage incurred during impact loading. It appears that negative curvatures inhibit delamination growth. Madan and Sutton 95 conducted low velocity impact tests on a number of stiffened panels designed for use in advanced wing structures. Their results indicated that the impact resistance of the structure was poor if

2. Conclusions
The impact resistance of a multidirectional laminate is strongly dependent upon the specific orientation of the plies. Unidirectional laminates should be avoided since
358 COMPOSITES. SEPTEMBER 1991

Impact energy = 1.8 J

Impact energy = 7.0 J

Length = 50 mm

Length = 50 mm

n D
Length = 75 mm Length = 100 mm

the material's behaviour as well as the structural response of the target. The latter has already been discussed in some detail above. In summary, it appears that the low velocity impact resistance of a composite is strongly dependent upon its ability to absorb energy elastically. Consequently, the strain energy absorbing capability of the fibres as well as the geometrical configuration of the target are of great importance. At very high rates of strain the structure responds in a local mode and the strain energy absorbing of the fibres and structure is less important. Here, the magnitude of the energy dissipated in mechanisms such as delamination, debonding and pull-out may become important. In recent years more and more attention has been given to determining the rate dependence of the ultimate properties of composites and their constituentsz~'24`26'28,~-gs. Harding and co-workers25'27'2s'96 have examined the strain rate sensitivity of Kevlar, GFRP and C'FRP. They have shown that carbon fibre composites are rate-insensitive when tested in fibre-dominated modes, whereas GFm' and Kevlar composites exhibit a distinct rate-dependent behaviour with modulus and tensile strength increasing with rate. Other workers have assessed the rate dependence of matrix-dominated modes of failure such as interlaminar fracture23'99. Such tests are particularly useful since delamination (a matrix-dominated mode of failure) is particularly detrimental to the compressive strength of a laminated composite ~'11. DCB tests on carbon fibre/epoxy composites t2 have shown that the Mode I interlaminar toughness does not vary with strain rate, Fig. 18. Similar tests on carbon fibre/PEEK (APC2), a thermoplastic matrix composite, have identified a distinct rate dependence :3, Fig. 18. Over a wide range of strain rates the Mode I fracture toughness remains invariant of strain rate. However, beyond a certain threshold, the toughness drops dramatically to approximately 20% of its original value. This gives cause for concern, suggesting that the impact resistance of this material may be poor at high rates of strain. Indeed, high velocity impact tests on APC2 have suggested that beyond a certain threshold velocity a change in failure mode o c c u r s 63 and the material experiences a sudden drop in mechanical performance59.
5O00 CF/PEEK 4000 CF/epoxy E"
I

Length = 75 mm

Length = 100 mm

U U
Length = 150 mm Length = 150 mm Fig, 17 Effect of beam l e n g t h o n d a m a g e d e v e l o p m e n t in a 16-ply (0,+1-45 ) CFRP composite subjected t o h i g h v e l o c i t y i m p a c t l o a d i n g (Ref 29)

the flanges were either too stiff or too flexible. This is clearly related to the structural stiffness effects shown previously in Fig. 13. For optimum impact resistance a compromise is again required.
2.

Conclusions

In order to design components and structures for impact resistance, geometrical effects need to be fully understood. In the case of low velocity impact loading, the size of the specimen or component is a critical parameter in determining its dynamic response. Here again the response of the target as well as the amount of damage incurred is related to the target's ability to store energy elastically. As a result of the lower level of transverse constraint, beams tend to be capable of absorbing more energy than larger structures such as circular plates. Care has to be taken, therefore, when using simple beam-like specimens to evaluate the dynamic response of more complex structures. High velocity impact loading by a light projectile induces a very localized form of target response, resulting in much of the incident energy being dissipated in a very small volume. Here, the areal geometry of the target is less important and simple coupons can often be used to characterize the response of the full-scale structure. Rate The rate at which the structure is loaded affects both

D Unstable Stable

3000

ca

D=
mo o o o

ta o oo

Q o o o o o

o o o o [] o o o oo o o

1000

+,, 10-3 (s -~ )

10-2

%
10 o

10-6

10-s

10-4

10-1

Fig. 18 Variation o f Gic w i t h s t r a i n rate f o r c a r b o n fibre/PEEK and a carbon fibre/epoxy composite (Refs 24 a n d 102)

COMPOSITES. SEPTEMBER 1991 359

These results suggest that care should be taken when using data obtained from static tests in order to characterize the dynamic response of composite materials. Brittle thermosets such as epoxy resins do not appear to exhibit a significant rate-dependent behaviour whereas tougher systems such as the latest generation of thermoplastic-based composites do 12. Clearly, simple static tests would fail to identify such strain rate sensitivity and may rank the materials incorrectly. SUMMAR Y AND CONCLUSIONS Concerns expressed regarding the impact resistance of fibre-reinforced composites are without doubt wellfounded. Low energy impacts are capable of generating large areas of delamination, resulting in significant reductions in residual strength. In recent years a concerted effort has been made by the materials manufacturers to improve the impact resistance and damage tolerance characteristics of continuous fibre composites. In this review some of the fundamental parameters governing the impact response and subsequent load-bearing properties of components manufactured from these materials have been identified. The following conclusions may be drawn from the present review. 1) The strain energy absorbing capacity of the fibres is one of the most important parameters in determining the impact resistance of a composite structure. Fibres that have a large area under the stress/strain curve tend to be better suited to energy absorbing applications. 2) The Mode II (forward shear) properties of the matrix appear to determine the level of damage incurred during impact and therefore the residual compressive properties of the composite. For good compression strength after impact the Mode II interlaminar fracture toughness should be high. 3) The strength of the fibre/matrix interphase region can be adapted to the required application. For example, if a projectile has to be stopped and residual properties are not important, then the interphase region should be weak to encourage failure through gross splitting and delamination. If damage containment is required, then the level of treatment applied to the surface of the fibres should be greater. 4) The fibre stacking sequence determines both the elastic energy absorbing capability of the composite as well as the failure mode. For damage containment, laminates with abrupt changes in fibre orientation (for example, one with +/-45 interfaces) should be avoided. 5) Geometrical effects are significant under conditions of low velocity impact loading. Varying the geometry changes the target's ability to store energy and therefore its impact resistance. Large targets are not necessarily better energy absorbers than small coupons. Care should be taken, therefore, when interpreting data from tests on laboratory-size specimens. 6) Varying the impact velocity and therefore the strain rate affects both the material's properties as well as
360 COMPOSITES. SEPTEMBER 1991

the target response. Low velocity impact loading by a heavy object induces an overall target response, whereas high velocity impact by a light projectile induces a localized mode of target deformation resulting in energy being dissipated over a small region immediate to the point of impact. In the latest generation of tough composites matrixdominated modes of fracture appear to show a distinct rate dependency and care should be taken when using static tests to characterize dynamic behaviour. ACKNOWLEDGEMENTS The authors would like to acknowledge the financial support of the National Science Foundation Science and Technology Center for High Performance Polymers, Adhesives and their Composites at Virginia Tech. This paper was originally submitted for 'Bonding and Repair of Composites II', Zurich, Switzerland, March 1991 (this conference was cancelled). REFERENCES
1 Mmders, P.W. and Harris, W.C. 'A parametric study of composite performance in compression-after-impact testing' SAMPE Journal 22 (1986) pp 47-51 2 Almy, M. 'Post damage capabilityofcarbon fibre reinforced matrices' Proc lnt Conf on Polymersfor Composites (The Plastics and Rubber Iost, London, UK, 1987) pp 11.1-11.10 3 Stellbrink, K.K.U. 'Improved impact damage tolerance' Proc European Syrup on DamageDevelopmentand FailureProcesses in CompositeMateriah edited by I. Verpoest and M. Wevers (Leuven, Belgium, 1987) 4 Curson, A.D., Leach, D.C. and Moore, D.R. 'Impact failure mechanisms in carbon fiber/PEEK composites' J Thermoplastic Composite Mater3 (1990) pp 24-31 5 Redmk, S. and Sna, C.T. 'Optimal use of adhesive layers in reducing impact damage in composite laminates' in Composite Structures, Vol 2, DamageAssessmentand MaterialEvaluation (Elsevier Applied Science Publ, 1987) pp 2.18-2.31 6 Hong, S. and IAu, D. 'On the relationship between impact energy and delamination area' Exptl Mech 13 (1989) pp 115-120 7 Boll, D.J., Bracero, W.D., Weidaer, J.C. and Man'i, W.J. 'A microscopy study of impact damage on epoxy-matrix carbon fibre composites' ProcInt Conf on Post FailureAnalysis Techniquesfor FiberReinforced Composites, OH, USA, 1985paper 8 8 Takeda, N., Sierakewski, R.L., Ross, C.A. and Malvera, L.E. 'Delamination-crack propagation in ballistically impacted glass/epoxy composite laminates' Expil Mech 4 (1980) pp 19-25 9 Takeda, N., Sierakowski, R.L. and Malvern, L.E. 'Transverse cracks in glass/epoxy cross-ply laminates impacted by projectiles' J MaterSci 16 (1981) pp 2D08-2011 10 Shadbelt, P.J., Cerraa, R.S.J. aad Ruiz, C. 'A preliminary investigation of plate perforation by projectiles in the subordnance range' Report No 1372/81 (Univ of Oxford, UK, 1981) 11 Bradshaw, F.J., Derey, G. and Skley, G.R. 'Impact resistance of carbon reinforced plastics' RAE TR 72240(MOD, 1972) 12 Avery, J.G. DesignManualfor Impact Damage TolerantAircraft Structures, AGARDograph No 238 (NATO, 1981) 13 Retem, A. 'Residual flexural strength of FRP composite specimens subjected to transverse impact loa"drag'SAMPEJourna124 No 2 (1988) pp 19-25 14 Rea, C.A. and Sierakowski, R.L. 'Studies on the impact resistance of composite plates' Composites4 (1973) pp 157-161 15 Broatman, L.J. gad Retem, A. 'Impact strength and toughness of fiber composite materials' in Foreign ObjectImpact Damage to Composites, ASTM STP 568 (American Society for Testing and Materials, 1975) pp 114--133 16 Helflmtim~,J.D. 'Charpy impact of unidirectional graphite/ aramid/epoxy hybrid composites' in CompositeMaterial~: Testingand Design (Fourth Conference),ASTM STP 617 (American Society for Testing and Materials, 1977)pp 375-388

17 Husmn, G.E., Whitney, J.M. and Halpin, J.C. 'Residual strength characterization of laminated composites subjected to impact loading' in Foreign Object Impact Damage to Composites op.cit, pp 92-113 18 Beaumont, P.W.R., Riewald, P.G. and Zweben, C. 'Methods for improving the impact resistance of composite materials' ibid. pp 134-158 19 Novak, R.C. and de Crescente, M.A. in Composite Materials: Testing and Design (Second Conference), ASTM STP 497 (American Society for Testing and Materials, 1972) pp 311-323 20 Cheresh, M.C. and McMichael, S. 'Instrumented impact test data interpretation' in Instrumented Impact Testing of Plastics and Composite Materials, ASTM STP 936 edited by S.L. Kessler, G.C. Adams, S.B. Driscoll and D.R. Ireland (American Society for Testing and Materials, 1987) pp 9-23 21 Bader, M.G. and Ellis, R.M. 'The effect of notches and specimen geometry on the pendulum impact strength of uniaxial CFRP' Composites 5 (1974) pp 253-258 22 Kakarala, S.N. and Roche, J.L. 'Experimental comparison of several impact test methods' in Instrumented Impact Testing of Plastics and Composite Materials op. cit. pp 144-162 23 GilllespieJr, J.W., Carlsson, L.A. and Smlley, A.J. 'Rate dependent Mode I interlaminar crack growth mechanisms in graphite/epoxy and graphite/PEEK' Composites Sci and Technol 28 (1987) pp 1-15 24 Beguefin, P. and Barbezat, M. 'Caract6risation m6canique des polym6res et composites ~tl'aide d'une machine d'essais rapides' Proc 5th Journ~e Nationale D YMA T, Bordeaux, France, 1989 25 Harding, J. 'Impact damage in composite materials' Sci and Engng of Composite Mater I (1989) pp 41--68 26 AmiJima, S. and Fujli, T. 'Compressive strength and fracture characteristics of fiber composites under impact loading' in Progress in Science and Engineering of Composites, Proc 4th lnt Confon Composite Mater, 1982 edited by T. Hayashi, K. Kawata and S. Umekawa (Okasan & Co Ltd, Tokyo) pp 399413 27 Bai, Y. and Harding, J. 'Fracture initiation in glass reinforced plastics under impact compression' in Structural Impact and Crashworthiness, Vol2 edited by J. Morton (Elsevier Applied Science Publishers, 1984) pp 482--493 28 Harding, J. and Welsh, L. 'A tensile testing technique for fibre.reinforced composites at impact rates of strain' J Mater Sci 18 (1983) pp 1810-1826 29 Cantwell, W.J. 'Impact damage in carbon fibre composites' PhD thesis (Univ of London, UK, 1985) 30 Rhodes, M.D., winlams, J.G. and Starnes Jr, J.H. 'LOwvelocity impact damage in graphite-fiber reinforced epoxy laminates' Proc 341h Ann Tech Conf, Society of the Plastics lnst, 1979section 20-D pp 1-10 31 Morton, J. manuscript in preparation 32 Degrleck, J. and Dechaene, R. 'Real time recording of transverse impact experiments on composite laminates' in Composites Evaluation, Proc 2nd lnt Conf on Testing, Evaluation and Quality Control of Composites (TEQ C 87) (Butterworths, London, UK, 1987) pp 61--68 33 Elber, W. 'Failure mechanics in low velocity impact on thin composite plates' NASA Technical Paper 2152 (1983) 34 Vedula, M. and Koezak, M.J. 'Impact resistance of cross-plied polyphenylene sulfide composites' Proc Fourth Japan-US Conf on Composite Materials (Technomic Publ Co Inc, 1988) pp 7281 35 Clark, G. 'Modelling of impact damage in composite laminates' Composites 20 (1989) pp 209-214 36 Adams, D.F. and Miller, A.K. 'An analysis of the impact behavior of hybrid composite materials' Mater Sci and Engng 19 (1975) pp 245-260 37 Hancox, N.L. 'Izod impact testing of carbon fibre reinforced plastics' Composites 2 (1971) pp 41-45 38 Chamls, C.C., Hanson, M.P. and Seraflni, T.T. 'Impact resistance of unidirectional fiber composites' in Composite Materials: Testing and Design (Second Conference) op.cit, pp 324-349 39 Chamis, C.C. and Sinclair, J.H. 'Impact resistance of fiber composites: energy absorbing mechanisms and environmental effects' in Recent Advances in Composites in the United States and Japan, ASTM STP 864 edited by J.R. Vinson and M. Taya (American Society for Testing and Materials, 1985) pp 326-345 40 Beaumont, P.W.R. 'Fracture mechanisms in fibrous composites' in Fracture Mechanics, Current Status, Future Prospects edited by R.A. Smith (Pergamon Press, 1979) pp 211-233

41 Morris, A.W.H. and Smith, R.S. 'Some aspects of the evaluation of the impact behaviour of low temperature fibre composites' Fibre Sci and Technoi 3 (1971) pp 219-242 42 Davies, P., Cant'well, W.J., Richard, H., MouHn, C. and Kanseh, H.H. 'Interlaminar fracture testing of carbon fibre/ PEEK composites validity and applications' in Developments in the Science and Technology of Composite Materials, Proc ECCM3, Bordeaux, 1989 edited by A.R. Bunseli, P. Lamicq and A. Massiah (Elsevier Applied Sci Publ) pp 747-755 43 Cantwell, W.J., Curtis, P.T. and Morton, J. 'An assessment of the impact performance of CFRP with high strain carbon fibres' Composites Sci and Techno125 (1986) pp 133-148 44 De~-ey,G., Skley, G.R. and H u t e h t ~ , J. 'Impact properties of carbon fibre/Kevlar 49 hybrid composites' Composites 9 (1978) pp 25-32 45 Lesser, D. and Leach, D. 'Compressive properties of thermoplastic matrix composites' Proc 341h lnt SAMPE Syrup, May 1989 pp 1464-1473 46 Cantweli, W.J., Curtis, P.T. and Morton, J. 'Low velocity impact damage tolerance in CFRP laminates containing woven and non-woven layers' Composites 14 (1983) pp 301-305 47 Humtoa, D.L. 'Composite interlaminar fracture: effect of matrix fracture energy' Composites Tech Review 6 (1984) pp 176-180 48 Beseem, W.D., Cottingham, R.L., Jones, R.L. and Peyser, P. 'The fracture of epoxy and elastomer modified epoxy polymers in bulk and as adhesives' J Appl Polym Sci 19 (1975) pp 25452562 49 Williams, J.G. and Rhodes, M.D. 'Effect of resin on impact damage tolerance of graphite/epoxy laminates' in Composite Materials: Testing and Design (Sixth Conference), ASTM STP 787 edited by I.M. Daniel (American Society for Testing and Materials, 1982) pp 450--480 50 Budmall, C.B. and Partridge, I.K. 'Phase separation in epoxy resins containing polyethersulphone' Polymer 24 (1983) pp 639644 51 Recker, H.G., AUspaeh, T., Alstadt, V., Folda, T., Heckman, W., lttemann, P., Teach, H. and Weber, T. 'Highly damage tolerant carbon fiber epoxy composites for primary aircraft applications' SAMPE Quarterly 21 No 1 (1989) pp 46-51 52 l)lamant, J. and Moulton, R.J. 'Development of resins for damage tolerant composites w a systematic approach' SAMPE Quarterly 16 No 1 (1984) pp 13-21 53 Bravene:, L.D., Filippov, A.G. and Dewhirst, K.C. 'New concepts in damage to tolerant composites-- lightly crosslinked thermosets' Proc 341h lnt SAMPE Syrup, May 1989pp 714--725 54 Sykes, G.F. and Steakley, D.M. 'Impact penetration studies of graphite/epoxy laminates' Proc 12th Nat SAMPE Tech Conf, Oct 1980 pp 482-493 55 Evam, R.E. and Masters, J.E. 'A new generation of epoxy composites for primary structural applications: materials and mechanics' in Toughened Composites, ASTM STP 937 edited by N. Johnston (American Society for Testing and Materials, 1987) pp 413-436 56 Masters, J.E. 'Correlation of impact and interleaf delamination resistance in intefleafed laminates' Proc Sixth lnt Conf on Composite Mater~Second European Conf on Composite Mater edited by F.L. Matthews, N.C.R. Buskell, J.M. Hodgkinson and J. Morton (Elsevier Appfied Science Publ, 1987) pp 3.96-3.107 57 Hlrschbuehler, K.R. 'A comparison of several mechanical tests used to evaluate the toughness of composites' in Toughened Composites op.cit, pp 61-73 58 Tobukure, K., Odagiri, N., Ite, Y. and Nishimura, T. 'High impact resistance CFRP from thermosetting system of Torayca Tg00H/3900 prepreg' in New Generation Materials and Processes edited by F. Saporiti, W. Merati and L. Peroni (SAMPE, 1988) pp 293-306 59 Dan-Jumbo, E., IAx~ood, A.R. and San, C.T. 'Impact damage characteristicsof bismaleimides and thermoplastic composite laminates' in Composite Materials: Fatigue and Fracture, 2nd Vol, ASTM STP 1012 edited by P.A. Lagace (American Society for Testing and Materials, 1989) pp 356-372 60 Leach, D.C. and Moore, D.R. 'Toughness of aromatic polymer composites reinforced with carbon fibres' Composites Sci and Techno125 (1985) pp 131-161 61 Davies, P., Cantwell, W., Moulin, C. and Kansch, H.H. Composites Sci and Techno136 (1989) pp 153-166 62 Davies, P., CamtweU,W.J. and Kamch, H.H. 'Measurement of initiation values of GIc in IM6/PEEK composites' Composites Sci and Techno135 (1989) pp 301-313

COMPOSITES. SEPTEMBER 1991 361

63 Morton, J. and Gedwtn, E.W. 'Impact response of tough carbon fibre composites' Composite Struct 13 (1989) pp 1-19 64 Moere, D.R. and Predlger, R.S. 'A study of low energy impact of continuous carbon fibre reinforced composites' presented at the 8th Ann NRCC/IMRI Syrup, Composites 87, Quebec, Canada, Nov 1987 65 Davies, P. 'Delamination behaviour of thermoplastic matrix composites' PhD thesis (University of Compiegne, France, 1987) (in French) 66 CantweU, W.J., Davies, P., Jar, P.-Y., Bourban, P.-E. and Kau~h, H.H. 'Joining and repair of carbon fibre PEEK composites' in Plastics-Metals-Ceramics edited by H.L. Horufeld (SAMPE, 1990)pp 411--426 67 Evans, R.E., Masters, J.E. and Courter, J.L. 'Toughened interleafed composites' in Advanced Composites (American Society for Metals, 1985) pp 249--257 68 Griffin, C.F. 'Damage tolerance of toughened resin graphite composites' in Toughened Composites op.cit, pp 23-33 69 Lchmmm, S., Megerdiglan, C. and Papalla, R. 'Carbon fiber/ resin matrix interphase: effect of carbon fiber surface treatment on composite performance' SAMPE Quarterly 16 No 3 (1985) pp 7-13 70 Drzal, L.T. and Rich, M.J. 'Effect of graphite fiber/epoxy matrix adhesion on composite fracture behavior' in Recent Advances in Composites in the United States and Japan op. cir. pp 16--26 71 Rogers, K.F., Sidey, G.R. and Kinstan-Lee, D.M. 'Ballistic impact resistance of carbon-fibre laminates' Composites 2 (1971) pp 237-241 72 Dorey, G. 'Relationship between impact resistance and fracture toughness in advanced composite materials' in Effect of Service Environment on Composite Materials, A GARD CP 288 (1980) 73 Bless, S.J. and Hartman, D.R. 'Ballistic penetration of S-2 glass laminates' Proc 2lst Int SAMPE Con[, Sept 1989 pp 852-866 74 McGarry, F.J., Manden, J.F. and Knwamoto, J. 'Impact resistance of rubber modified carbon fiber composite' in Advanced Composites op.cit, pp 195-205 75 Dorey, G. 'Damage tolerance and damage assessment in advanced composites' in Advanced Composites edited by I.K. Partridge (Elsevier Applied Science Publ, 1989) chapter 11 76 Dorey, G. 'Fracture bebaviour and residual strength of carbon fibre composites subjected to impact loads' in Failure Modes of Composite Materials with Organic Matrices and Their Consequences in Design, AGARD CP 163 (1975) paper 8 77 Ikewicz, L.B., i)est, E.F. and Coggeshall, R.L. 'A model for compression after impact strength evaluation' Proc 21st lnt SAMPE Tech Con[, Sept 1989pp 130-140 78 Kant, C.Y. and Walker, J.V. 'Toughened composites selection criteria' in Toughened Composites op.cit, pp 9-22 79 Demuts, E. and Sharpe, P. 'Tougher advanced composite structures' 28th Structures, Structural Dynamics and Materials Con[, Monterey, CA, USA, April I987 (AIAAIASMEIASCEI AHS) pp 385-393 80 Stevanovle, M., Kostic, M., Stecenko, T. and Briski, D. 'Impact behaviour of CFRP composites of different stacking geometry' in Composites Evaluation, Proc TEQ C 87 op.cit, pp 78-83 81 Dorey, G. 'Fracture of composites and damage tolerance' in Practical Considerations of Design, Fabrication and Tests for Composite Materials, A GARD Lecture Series 124 (1982) 82 Nolet, S.C. and Sandusky, P.M. 'Impact resistant hybrid composite for aircraft leading edges' SAMPE Quarterly 17 No 4 (1986) pp 46-53 83 CantweH, W.J. and Morton, J. 'Geometrical effects in the low velocity impact response of CFRP' Composite Struct 12 (1989) pp 35-59 84 Liu, D. 'Impact-induced delamination - - a view of bending stiffness mismatching' J Composite Mater 22 (1988) pp 674--692 85 Vednla, M. and Keczak, M.J. 'Impact resistance of cross-plied polyphenylene sulfide composites' J Thermoplastic Composite Mater2 (1989) pp 154-163 86 Malllek, P.K. and Broutman, L.J. 'Static and impact properties of laminated hybrid composites' J Testing and Evaluation 5 (1977) pp 190-200 87 Hancex, N.L. and Wells, H. 'Izod impact properties of carbon fibre/glass fibre sandwich structures' Composites 4 (1973) pp 26-29 88 Adams, D.F. and Perry, J.L. 'Static and impact behaviour of graphite/epoxy composite laminates containing third-phase reinforcement materials' J Testing and Evaluation $ (1977) pp 114-123

89 Ko, F.K. and Hartnmn, D. 'Impact behavior of 2-D and 3-D glass/epoxycomposites' SAMPE Journal 22 No 4 (1986) pp 263O 90 Herrlek, J.W. 'Impact resistant multidimensional composites' Proc 12th Nat SAMPE Tech Con[, Oct 1980 pp 845-856 91 Madan, R.C. ' G - - a measure of damage tolerance of composites' Proc 20th lnt SAMPE Tech Con[, Sept 1988 pp 403--412 92 Dow, M.B, and Smith, D.L. 'Damage-tolerant composite materials produced by stitching carbon fabrics' 21st Int SAMPE Tech Con[, Sept 1989 93 Su, K.B. 'Delamination resistance of stitched thermoplastic matrix composite laminates' in Advances in Thermoplastic Matrix Composite Materials, ASTM STP 1044 edited by G.M. Newaz (American Society for Testing and Materials, 1989) pp 279--300 93 Cantwell, W.J. 'The influence of target geometry on the high velocity impact response of CFRP' Composite Struct 10 (1988) pp 247-265 94 Ganse, L.W., Resenfeld, M.S. and Vining, R.E. 'Effect of impact damage on the XFV-12A composite wing box' Proc 25th Nat SAMPE Conf, 1980 pp 679--690 95 Madan, R.C. and Sutton, J.O. 'Design, testing and damage tolerance study of bonded stiffened composite wing cover panels' 29th Structures, Structural Dynamics and Materials Con[, Williamsburg, VA, USA, April 1988 (AIAA/ASME/ASCE/ AHS) pp 623--630 96 Welsh, L.M. and Harding, J. 'Effect of strain-rate on the tensile failure of woven-reinforced polyester resin composites' Report OUEL 1578/85 (Univ of Oxford, UK, 1985) 97 Daniel, I.M., Yaniv, G. and Anser, J.W. 'Rate effects on delamination fracture toughness of graphite/epoxy composites' in Composite Structures, Vol2 op. cit. pp 2.258-2.272 98 Karger-Kocsis, J. and Frledrkh, K. 'Temperature and strainrate effects on the fracture toughness of PEEK and its short glass-fibre reinforced composite' Polymer27 (1986) pp 17531760 99 Mall, S., Law, G.E. and Katea~i=n, M. 'Loading rate effect on interlaminar fracture toughness of a thermoplastic composite' J Composite Mater 21 (1987) pp 569-579 100 Starnes, J.H., Rhodes, M.D. and Williams, J.G. 'Effect of impact damage and holes on the compressive strength of a graphite/epoxy laminate' in Nondestructive Evaluation and Flaw Criticalityfor Composite Materials, ASTM STP 696 edited by R.B. Pipes (American Society for Testing and Materials, 1979) pp 145-171 I01 Sarma Avva, V. and Padmanabha, H.L. 'Compressive residual strength prediction in fiber-reinforced laminated composites subjected to impact loads' in Advances in Fracture Research, Proc 6th lnt Conf on Fracture edited by S.R. Valluri, D.M.R. Taplin, P. Rama Rao, J.F. Knott and R. Dubey (1984) pp 28972907 102 Barbezat, M. 'The influence of loading rate on the behaviour of epoxy composites' PhD thesis (Ecole Polytechnique F~d~rale de Lausanne, Switzerland, 1990) (in French) 103 Smihy, A.J. 'Rate sensitivity of interlaminar fracture toughness in composite materials' M.S. thesis (Univ of Delaware, USA. 19S5) 104 Harris, B., Beaumont, P.W.R. and Monennin de Ferran, E. 'Strength and fracture toughness of carbon fibre polyester composites' J Mater Sci6 (1971) pp 238-251 105 Bal~lyopadhyay, S., Gellert, E.P., SHva, V.M. and Underweod, J.H. 'Microscopic aspects of failure and fracture toughness in advanced composite materials' J Composite Mater 23 (1989) pp 1216-1231 106 Kirk, J.N., Mnnro, M. and Beaumont, P.W.R. 'The fracture energy of hybrid carbon fibre and glass fibre composites' J Mater Sci 13 (1978) pp 2197-2204

AUTHORS
W . J . C a n t w e l l , w h o is with t h e L a b o r a t o i r e d e P o l y m e r e s , E c o l e P o l y t e c h n i c F t d t r a l e d e L a u s a n n e , 1007 L a u s a n n e , S w i t z e r l a n d , is c u r r e n t l y a visiting scientist at V i r g i n i a P o l y t e c h n i c I n s t i t u t e a n d S t a t e U n i v e r s i t y . J. M o r t o n is w i t h t h e D e p a r t m e n t o f E n g i n e e r i n g Science and Mechanics, Virginia Polytechnic Institute a n d S t a t e U n i v e r s i t y , B l a c k s b u r g , V A 24061--0219, USA.

362

COMPOSITES. SEPTEMBER 1991

Das könnte Ihnen auch gefallen