impacto baja velocidad

8/18/2019 Impacto Baja Velocidad

1/9

S1359-835X( N)OOW4-7

Composites Purr A 27A (1996) 112331131

Copyright 0 1996 Elsevier Science Limited

Printed in Great Britain. All rights reserved

1359-835X/96/ 15.00

ELSEVIER

Review

Review of low-velocity impact properties

of composite materials

M. 0. W. Richardson” and M. J. Wisheart

Institute of Polymer Technology and Materials Engineering, Loughborough University of

Technology, Loughborough, Leicestershire LEll 3TU, UK

Received 28 October 1994; revised 19 April 1996)

This paper is a review of low-velocity impact responses of composite materials. First the term ‘low-velocity

impact’ is defined and major impact-induced dama ge mo des are described from onset of dama ge throu gh to

final failure. Then, the effects of the composite ’s constituents on impact properties are discussed and post-

impact perform ance is assess ed in terms of residual stre ngth. Copyright 0 1996 Elsevier Science Limited

(Keywords:

ow-velocity impact; damage modes; constituent properties; specimen geometry; post-impact residual strength;

review)

INTRODUCTION

This review ha s been carried out as part of a research

program me investigating the application of pultruded

composite systemsle3

to the construction of freight con-

tainers. This is a realistic goal, ow ing to opportunities

which have arisen in the container industry&I6 and tech-

nological advanc es in the putrusion process’7-‘9. Con-

tainers have traditionally been metal constructions, and

extensive rese arch h as been performed on the impa ct

response of meta ls over a wide range of velocities. Impac t

damage in metals is easily detected as damage starts at

the impacted surface; however, damage in composites

often begin s on the non-imp acted surface or in the form

of an internal delam ination.

Impact damage is generally not considered to be a

threat in metal struc tures becau se, ow ing to the ductile

nature of the material, large amo unts of energy may be

absorbed . At yield stress the mate rial may flow for very

large strains (up to 20% ) at constant yield before work

hardening. In contrast, composites can fail in a wide

variety of mode s and contain barely visible impa ct

damage (BVID) which nevertheless severely reduces the

structural integrity of the comp onent. Mo st comp osites

are brittle and so can only absorb energy in elastic

deformation and through dam age mechanisms, a nd not

via plastic deformation. The term damage res is tance

refers to the amount of impact damage which is induced

*

To

whom correspondence should be addressed

in a composite system. Clearly, the vast majority of

impacts on a composite plate will be in

the

transverse

direction

but due to the lack of through-thick ness rein-

forcement, transverse dam age resistance is particularly

poor. Interlamina r stresses-she ar and tension-are

often the stresses that cause first failure due to the

correspondingly low interlaminar strengths. A s a result,

design failure strains of 0.5% are used to guard against

impac t failure, resulting in a failure to take advanta ge of

the excellent in-plane strength and stiffness properties of

composites.

DEFINITION OF LOW-VELOCITY IMPACT

Generally, impa cts are categorized into either low or

high velocity (and sometim es hyper velocity), but there is

not a clear transition between c ategories and authors

disagree on their definition.

Sjijblom et aL2’ and Shivakumar and co-workers2’

define low-velocity impa ct as events whic h can be treated

as quasi-static , the upper limit of which can vary from

one to tens of ms-’ depen ding on the target stiffness,

material properties and the impacto r’s ma ss and stiff-

ness. High-velocity impac t response is dom inated by

stress wave propagation through the material, in which

the structure does not have time to respond, leading to

very localized damag e. Boun dary condition effects can

be ignored because the impac t event is over before the

stress waves have reached the edge of the structure. In

1123


2/9

Re v i e w o f l o w v e l o c i t y im p a c t p r o p e r t i e s : M . 0 . W . R i c h a r d s o n a n d M . J . W i s h e a r t

low-velocity impa ct, the dynam ic structural response

of the target is of utmost importance as the contact

duration is long enough for the entire structure to

respond to the impact and in consequence more energy is

absorbed elastically.

Cantwell and Morton2’

conveniently classified low

velocity as up to 10 m s-i, by considering the test tech-

niques which are generally employed in simulating the

impact event (instrumented falling weight impact testing

(IFWIT), Charpy, Izod, etc.) whilst, in contrast, Abrate2 3

in his review of impact on laminated composites stated

that low-velocity impa cts occur for impac t spe eds of less

than lOOms_‘.

Liu and M alvern24 and Joshi an d S un25 suggest that

the type of impac t ca n be classed ac cording to the

damage incurred, especially if damage is the prime con-

cern. High velocity is thus characterized by penetration-

induced fibre breakag e, and low velocity by delam ination

and matrix cracking.

Davies and Robinson26*27

define a low-velocity impac t

as being one in which the through-thickness stress wave

plays no significant part in the stress distribution and

suggest a simple model to give the transition to high

velocity. A cylindrical zone under the impac tor is con-

sidered to undergo a uniform strain as the stress wave

propagates through the plate, giving the compressive

strain as26

E, =

impact velocity

speed of sound in the material

For failure strains between 0.5 and l%, this gives the

transition to stress wave dom inated events at lo-

20 m s-* for epoxy comp osites.

MODE S OF FAILURE IN LOW-VELOCITY

IMPACT

The heterogeneous and anisotropic nature of fibre-

reinforced plastic (FRP) lamin ates gives rise to four major

mode s of failure (although many others could b e cited):

1) ma t r i x mode -cracking occurs parallel to the fibres

due to tension, comp ression or shear;

2)

de l am i n a t i o n

mode-produ ced by interlamina r stresses;

3)

j i b r e

mode -in-tension fibre breakage and in-

compression fibre buckling; and

Ii _11

mmx

J--k

._

(a) uansvcrse view

impacted O/90/0 composite plate

4) pene t r a t i on - t he impactor completely perforates the

impacted surface.

It is very important to identify the mode of failure

because this will yield information not only about the

impact event, but also regarding the structure’s residual

strength. Interaction between failure mode s is also very

important in understanding damage mode initiation an d

propagation24.

M a t r i x damage

The majority of the impa ct test work reported in the

literature has involved low-energy testing, in the range of

1 to 5 J approximately (i.e. that which ca uses only mini-

mal damage). It is this work that has revealed informa-

tion about m atrix cracking and delamination initiation.

Ma trix dam age is the first type of failure induced by

transverse low-velocity impac t, and usually takes the

form of matrix cracking but also debonding between

fibre and matrix. Matrix cracks occur due to property

mismatching between the fibre and matrix, an d are usu-

ally oriented in planes p arallel to the fibre direction in

unidirectional layers. Joshi and Sun2’ reported a typical

crack and delamination pattern show n in

Figure I .

The matrix cracks in the upper layers Figure la ) and

the middle layer Figure lb ) start under the edges of the

impactor. These shear cracks29 are formed by the very

high transverse shear stress through the material, and are

inclined at approximately 45”. The transverse shear

stresses are related to the contact force and contact area.

The crack on the bottom layer of

F igu re la

is termed a

bend ing crack because it is induced by high tensile bend-

ing stresses and is characteristically vertical. The bending

stress is closely related to the flexural deforma tion of the

laminate.30 Lee and Sun31 reached the same conclusions

in their analyses. Cantwell and Morton32 em phasized

that the type of matrix cracking which oc curs is depend-

ent on the global structure of the impac ted specim ens.

For long thin specimens bending cracks in the lower

layers occur due to excessive transverse deflection a nd

subsequent m embrane effects predominate, whereas short

thick spe cimen s are stiffer and so higher p eak contact

forces induce transverse shear cracks under the impactor

in the upper plies.

Liu and M alvern24 presented a detailed view of matrix

cracking which agreed with the above, w hilst Wu and

delamination

(b) longitudinal view

Figure Initial damage

in an

1124


3/9

Review of low-velocity impact properties: M. 0. W. Richardson and M. J. Wisheart

Springer3” reported detailed loca tions of matrix crack ing

for graphite/epoxy plates of various sta cking sequen ces.

Chang and co-workers29,34-38 have performed much

research in this area, and postulated that the bending

crack in the 90” layer is caused by a comb ination of (Tag,

gIl and g13 (Figure 2) for line-loading impact damage.

Their analysis also concluded that 033 was very small

relative to gll and ~7~~hroughout the impact event, and

that there is a critical energy below which no dama ge

occurs.

Delamination

A delamination is a crack which runs in the resin-rich

area (approximately 0.0007 mm in graphite/epoxy lami-

nates39 ) between plies of different fibre orientation and

not between lamina in the same ply group33,40>41.

Liu and Malvern24

compiled detailed connections

between delaminations and the areas over which matrix

cracks w ere found, for various lay-ups. Liu4* explained

that delamination was a result of the bending stiffness

mism atch between adjacen t layers, i.e. the different fibre

orientations between the layers. In his experime ntal work

he found that delamina tion areas were generally oblong-

shaped with their major axis being coincident with the

fibre orientation of the layer below the interface. For

O/90 laminates the shape became that of a distinct

peanu t. These results have been widely reported else-

where in the literature28,33,34,4346.He also stated that it is

the bending-induced stresses which are the major cause

of delamination, as both experiment and analysis

revealed that along the fibre direction the plate tends to

bend conca ve, whilst the bend is convex in the transverse

direction. Liu defined a bending mism atch coefficient

between the two adjacent laminates which includes

bending stiffness terms and predicts the peanut shape

reported for O/90 lamin ates. The greater the mism atch

(O/90 is the worst-case fibre orientation), the greater the

delam ination area will be. This is also affected by

material properties, stacking sequence and laminate

thickness4’.

Dorey48-50 has worked widely in this field and pro-

vides a simple expression for the elastic strain energy, E,

absorbed at the point of delam ination failure, which

suggests that this damage mode is more likely to occur

for short spans and thick laminates with low interlami-

nar shear strength.

21-*WL3

Energy = 9Et

f

where t = thickness, 7 = interlaminar shear strength

(ILSS), u’ = w idth, L = unsupported length and Ef =

flexural modulus.

Delamination initiation an d interaction with matri.x

cracking.

Delamination caused by transverse impact

only occurs after a threshold energy has been reached

and it has been observed that delamination only occurs

in the presence of a matrix crack35. Much detailed

work h as been performed to verify this and explain the

stress states which could cause this interaction.

Takeda et aL51 revealed for the first time the asso-

ciation between matrix cracking and delamination, and

showed that delaminations do not always run precisely in

the interface region, but can run slightly either side . Joshi

and Sun’* studied the delamination-matrix crack inter-

action for O/90/0 lamin ates subject to transverse point

impact. They concluded that when the inclined shear

crack in the upper layer

Figure la)

reaches the interface

it is halted (by the change in orientation of the fibres) and

so propagates between the layers as a delamination. This

delam ination is generally constrained by the middle

transverse crack

Figure lb).

The vertical bending crack

Figure lb)

is thought to initiate the lower interface

delamination, the growth of which is not constrained.

Matrix cracks which lead to delamination are known as

critical matrix cracks34.

Chang and colleagues29.34’35.52 erformed a series of

line-loading, low-velocity im pact tests and reported a

typical damage pattern for a O/90/0 composite as shown

in Figure 3. Change and Choi34 simulated these matrix

cracks in their three-dime nsional finite elemen t analysis

to study the stress in the vicinity of the cracks. They

concluded that delamination was initiated as a mode I

fracture process due to very high out-of-plane norma l

stresses caused by the presence of the matrix cracks and

Figure 2 Diagram of the stress components contributing to a bending

matrix crack in a transverse layer

matr i xcrack

\

clamped boundary

Figure 3

Typical matrix crack and delamination pattern from line-

load impact on a O/90/0 composite

1125


4/9

Revi ew of ow -vel oci ty impact propert i es: M. 0. W . Ri chardson and M . J. W ishearl

high interlamin ar shear stresses along the interface. In his

review on delamination, Gargs3 proposed that matrix

crack initiated delamination was due to the development

of the interlaminar normal and shear stresses at the

interfaces.

Liu and co-workers52 created an analytical model to

study the interaction of damage mechanisms due to line-

load impact, utilizing a fracture mechanics approach.

They showed that both bending cracks and shear cracks

could initiate delam ination, but that delamination

induced by shear cracks is unstable and that bending

crack induced de laminations grow in a stable manner

and proportional to the applied load.

Finn and Springer47’54

described in detail the stresses

which they believed cause impact-induced delamination.

All the modes w hich could be induced by impact-

bending, twisting and transverse shear-were con-

sidered, as were the restraints on the affected ply due

to layers above an d below. They concluded that if the

cracked ply group is above the interface, then (if the

upper interface of the ply group is unrestrained) cl2 and

23 contribute to delam ination; if the cracked ply group

is below the interface, ~7~~ nd g1 2 contribute to delami-

nation as long as the ply group lower interface is

unrestrained.

Mo st fracture m echanics analyses of the initiation and

growth of delam ination are difficult to apply becau se

they assum e a n initial flaw or crack size5’; however

Davies and Robinson27,

in a highly simplified isotropic

axisymm etric analysis for the threshold force for growth

of an internal circular d elamin ation in the mid-plan e,

show (surprisingly) that mode II strain energy release

rate is independ ent of delam ination radius. Therefore an

initial flaw size is not required and the threshold force is

given by:

p2 = gr2Eh3Guc

c

9(1 - z/2)

where

PC

threshold load, GIIc = critical strain energy

release rate, Y = Po isson’s ratio, h = plate thickness and

E = modulus.

The predictions from this equation for delam ination

initiation agreed well with their experime ntal data on

quasi-isotropic laminates.

Delami nation growt h.

Choi and Chang45 reported

that delamination growth was governed by interlaminar

longitudina l shear stress (c13) and transverse in-plane

stress (022) in the layer below the delam inated interface

and by the interlaminar transverse shear stress (~7~4 n

the layer above the interface.

Several investigators have introduced artificial delam i-

nations by including a thin foil between plies in the

manufacturing stage to assess delamination growth

from a known initial size56. Doxsee et a1.57 alculated

the energy abso rbed per unit area of delam ination

growth and found that this was constant (595 Jmp2).

Jih and Sun3’ concluded that the interlaminar fracture

1126

toughness was independent of delamination size and

that delamination area could be predicted from peak

impact force generated. Wu and Shyu4 also found that

there was a linear relationship between the peak force

and delamination area and, by extrapolating from the

results, they found a threshold force value for the onset

of delamination.

In their num erical simula tion of impac t-induced

delamination growth, Razi and Kobaysh?* concluded

that mode II was the dominant failure mode for propa-

gation, a view also put forward by Guild et aZ .43.

Fibre ail ure

This damage mode generally occurs much later in the

fracture process than matrix cracking and delamination,

and, as research has concentrated on the low-energy

mode s of dam age, there is less information on this area.

Fibre failure oc curs under the impactor due to locally

high stresses and indentation effects (mainly governed by

shear forces) and on the non-imp acted face due to high

bending stresses. Fibre failure is a precursor to cata-

strophic penetration mode . A simple equation for the

energy req uired for fibre failure due to back su rface

flexure is given by Dorey49 a s:

a2wtL

Energy = 18E

f

where 0 = flexural strength, Ef = flexural modulus,

w = width, L = unsupported length and t = specimen

thickness.

Penetration

Penetration is a macrosco pic mode of failure and

occurs when the fibre failure reach es a critical extent,

enabling the impac tor to completely penetrate the

materia159. Research into penetration impact has mainly

concentrated on the ballistic range; however, some low-

velocity impact work has been performed. Cantwell and

Morton32 showed that the impact energy penetration

threshold rises rapidly w ith ‘specimen thickn ess for

carbon fibre-reinforced plastic (CFR P). They also ana-

lysed the penetration process to calculate the energy

absorbed by shear-out (i.e. removal of shear plug),

delam ination and elastic flexure. This simplified analysis

predicted shear-out as the major form of energy absorp-

tion (50-60% depending on plate thickness).

El-Hab ak6’ tested a variety of glass fibre-reinforced

plastic (GFRP) composites at penetration loads and

concluded that the glass fibre treatment played a key

role in determin ing the perforation load. Whilst the

matrix had little effect, polyester wa s preferable to epoxy.

Dorey49 provided a very simplified analytical mode l of

penetration to give the energy absorbed as:

Energy = qtd

where y = fracture energy, d = diamete r of impacto r,

and t = plate thickness.


5/9


Damage m odes in randomly orientedfibre laminates

Most of the work reported above w as on laminates

consisting of unidirectional plies with varying fibre

orientation. In layers in which the fibres are unidirec-

tional it is quite straightforward to predict the orienta-

tion of matrix cracking. W hen the fibres are oriented

randomly, then crack patterns are less easy to establish .

Sheet moulding compound (SMC ) panels and continu-

ous filament mats (CF M) used in pultrusions are common

exam ples of randomly oriented short and long fibre

layers, respectively.

Clearly, a different approac h to defining damage

mode s is required for these composites. In their research

on SMC panels, Liu and Malvern24 found that matrix

cracks on the impacted surface were short and formed a

series of rings away from the point of contact, and

deduced that these were caused by the tensile strain wave

moving out from the centre of impac t. Both Chaturve di

and Sierakowski6’ and Khetan and Chang62 performed

work on glass/polyester SM C panels with air-gun equip-

men t (i.e. high velocity). W hilst the latter au thors

suggested that damag e could be quantified by a ‘damage

area’, the former au thors concluded from tensile residual

strength tests that more information was required on

failure mo des to be able to predict stiffness and strength

degradation. Liu

et a1.63,

in their work on the repair-

ability of SM C comp osites for the automotive industry,

defined three types of impac t-induced dam age: (1)

indentation (crushing of matrix under the impactor),

(2) bending fracture, and (3) perforation (i.e. dam age

resulting from penetration and associate d fracture).

INFLUENCE OF CONSTITUENTS ON THE

IMPACT RESPONSE OF COMPOSITE

MATERIALS

A fibre-reinforced compo site consists of two major con-

stituents (fibre and matrix) an d the interphase region,

which is the area of bond between fibre and m atrix.

The properties of each of these constituents affect the

threshold energies or stresses required to initiate the

different failure modes induc ed by impac t.

Fibres

Fibres are the main load-bearing constituent, provid-

ing the comp osite with the majority of its strength and

stiffness. The most comm on fibres are glass, carbon and

Kevlar. Carbo n is widely used in the aircraft industry

and in many structural applications as it has the highest

strength and stiffness values; how ever, it also is the most

brittle, with a strain to failure of 0.5 to 2.4%. Glass fibres

have a lower strength and stiffness but have a higher

strain to failure (~3.2%~ ) and are less expensive than

carbon fibres. The mec hanical properties of Kevlar lie

between those of carbon and glas?. Carbon’s design

ultimate allowab le strain is only 0.4% currently, whilst

improvements in damage tolerance performance would

allow a 50% improvement on this@‘.Thus a great deal of

the fibre’s superior performance characteristics cannot

be taken advantage of owing to its weakness with respect

to impact.

For resistance to low-velocity impa ct, the ability

to store energy elastically in the fibres is the fundam ental

parameter22.

This corresponds to the area under the

stress-strain curve, which is dictated by the fibre

modulus and failure strain. E-glass can therefore

absorb approxima tely three times the elastic energy of

carbon. Hybrid comp osites are often formed by adding

glass or Kevlar55>65

o carbon comp osites to improve

impact resistance, but moduli mismatching between

fibres increases the complexity of the design of hybrids.

Strain rate sensitivity of glass jibres.

There is con-

flicting information in the literature regarding the strain

rate sensitivity of glass fibres. In general, carbon fibres

are thought of as not being strain rate dependent20’48’67%68

and glass fibres as having a modulus and stiffness which

increase with strain rate48,68s70.Howeve r, in their review

in 1983, Sierakowski and Chaturvedi” concluded that

there was not enough information available to assess

the role of rate sensitivity of comp osite systems. That is

still the case today.

In their impact tests from 1 to 5.5ms-‘, Caprino

et a1.72

reported no strain rate effects for glass cloth/

polyester. However, over a wider strain rate range,

Sim s73 reported increasing flexural strength s for a glass

mat/polyester lamin ate over a range of 10e6 to 10-l m s-’

displacem ent rate for Charpy testing of com posites.

Hayes and Adam s74 constructed a specialized pendu-

lum impactor to study tensile strain rate effects, as

impact speeds increased from 2.7 to 4.9 m

s-’

for glass/

epoxy. They also performed static tensile tests. The

elastic modulus and strength in general increased w ith

impact velocity, but the trend was not consistent through-

out the dynamic range, an d the values at static loading

did not support this trend. In contrast to the belief that

carbon fibres are non-rate depend ent, they reported that

graphite/epoxy’s modulus decreased w ith impact speed

and at dynamic loads the ultimate strength and energy to

ultimate stress were lower than the static values.

Li et af.75 reported an increase in the tensile and

comp ressive strength and stiffness for glass from quasi-

static rates of strain (0.001 to 10~~‘) to high-velocity

impa ct (350 to llOOs_‘). They also noted through-

thicknes s strength increases for glass weave. How ever,

low-velocity impa ct indu ces stra in rates which lie in a

muc h narrower range than the work of Li; therefore, in

general, for low-velocity impac t testing fibre strain rate

effects can be ignored.

Matrix

In an FRP the polymeric matrix (usually a thermoset)

provides several key functions: it transfers the load to the

fibres, protects the fibres from damagin g themselv es an d

1127


6/9


aligns/stab ilizes the fibres. The majority of structural

application s employ epoxy resins as they meet the hot/

wet comp ressive strength requirements. However, epoxy

is brittle an d has poor resistance to crack growth (i.e.

delamination). Attempts to reduce matrix damage and

improve the interlamina r fracture toug hness of thermo-

set resins has involved incorporating plasticizing modi-

fiers, or adding rubbe r or thermop lastic particles to

the resin66. Howeve r, increa sed interlam inar fracture

toughness invariably reduces mechanical properties and

improvements made to the pure matrix are never trans-

ferred fully to the comp osite due to the presence of the

brittle fibres, which prevent growth of plastic zo nes in the

matrixz2. The inclusion of a thin discrete layer of very

tough, high shear strain resin can also be employed to

minimize delamination76.

The use of thermop lastic resins [e.g. poly(ether ether

ketone)] c an give an order of mag nitude increase

in fracture toughn ess over thermoset comp osites. Low

thermal stability and chem ical resistance, poor fibre-

matrix interfacial bonding and creep problem s have his-

torically prevented the use of thermop lastic compo-

sites66. The need for new production techniques still

holds back the use of thermoplastics, but as these

problems are overcome so thermoplastic-based compo-

site systems become more com petitive.

In t e r p h a s e r e g i o n

The interphase region betwee n fibre and matrix is of

vital importanc e. Usually, the surfaces of carbon fibres

are treated with an oxidative process in order to improve

the level of adhesio n between fibre and matrix, whilst

glass fibres are treated with a coupling agent. The inter-

phase region can affect the failure mode which occurs at

a given load; i.e. poor adhesio n results in failure a t low

transverse stress, leaving clean fibres. The bond strength

can be manipulated to improve the toughness by absorb-

ing energy in fibre-matrix debond ; ho wever, this reduces

the mech anical properties.

IMPACT PERFORMANCE OF COMPLEX

GEOMETRY SPECIMENS

The vast majority of impa ct te sting has been performed

on flat coupons in beam or plate format, either clamp ed

or simply supported. By simplifying the geometry, struc-

tural effects are minim ized and more information can be

gleaned about the material behaviour. However, many

composite components have a complex geometry and the

response of stiffened panels is a particularly importan t

area.

Dorey4* reported that the energy to cause BV ID

dropped significantly near the stiffeners, where the struc-

ture was less compliant, and that the stiffeners caused

damage to spread asymm etrically, as would be expected

over an area of non-uniform stiffness. Davie s and co-

authors27946 tated that impact forces will be higher in the

1128

stiffened regions, but that reduced deflections may lead

to sma ller strains and therefore less strain-induced

failure. At the edge of the stiffeners delam inations were

formed, whilst impacts directly over the stiffener caused

debonding between plate and stiffener. The damage

tended to extend down the stiffener (because the induced

forces will follow the stiffest path), which wo uld hav e

disastrous effects for a comp ression-loaded panel. Crater-

ing also occurred due to the very high forces induced in

the stiffened regions. Due to fear of stiffener-panel

debond, many manufacturers are using mechanical join-

ing technique s to avoid this problem, indicating that it is

an area of some concern.

POST-IMPACT RESIDUAL STRENGTH

As stated previously, due to the susceptibility of com-

posite materials to impact damage, dramatic loss in

residual strength and structural integrity results. The

term d ama g e t o l e ra n c e refers to a system’s ability to

perform post-impact. Even BVID can cause strength

reductions of up to 50%. Re sidual strengths in tension,

compression, bending an d fatigue will be reduced to

varying degrees depending on the dominant damage

mode.

Re s i d u a l t e n s i l e s t r e n g t h

Resid ual tensile strength2 3 normally follows a curve as

shown in

F i g u r e 4 .

In region I, no damage occurs as the

impac t energy is below the threshold value for dam age

initiation. Once the threshold has been reached, the

residual tensile strength reduces quickly to a minimum in

region II as the extent of dam age increases. Region III

sees a constant value of residual strength becaus e the

impact velocity has reached a point w here clean

perforation occurs, leaving a neat hole. In this region

the tensile residual strength can be estimate d by con-

sidering the dama ge to be equivalent to a hole the size of

the impactor. The minimum in region II is less than the

constant value in region III because the damage spreads

over a larger area tha n is produced at a higher velocity

when the dam age is more localized (resulting in a cleaner

hole)48. As the fibres carry the majority of tensile load in

the longitudina l direction, fibre dam age is the critical

Residual

Tensile

snength1\_.

II : II

III

Y

Impact Energy

Figure 4 Characteristic residual strength

v rsus

impact energy curve


7/9


damag e mo de. C aprino68 developed a linear elastic frac-

ture mechanics model to predict residual tensile strength

as a function of impa cting k inetic energy which gav e

good correlation to experimental results.

Residual compressive strength

Poor post-impact compressive strength (PICS) is the

greatest weakness of composite laminates in terms of

residual properties. This is mainly due to local instability

resulting from delamination causing larg e reductions in

compressive strength48’n.

As delamination can be pro-

duced by low-energy impacts, large strength reductions

in compression can occur for BVID . D elamination

divides the laminate into sub-laminates which have a

lower bending stiffness than the original laminate and are

less resistant to buckling loads&. Under a compressive

load, a delamination catl cause buckling in one of three

mod es23: global instability/buckling of the lamina te,

local instability (buckling of the thinner sub-laminate),

or a combination of the above. The mod e of failure

generally changes from global, to local, to mixed mode as

the delamination length increases .

lOm s_’ which are ordinarily introduced in the labora-

tory by mechanical test machines such as the IFWIT

technique. The contact period is such that the whole

structure has time to respond to the loading. The modes

of impact damage induced range from matrix cracking

and delam ination through to fibre failure and penetra-

tion. Damage mode interaction must also be understood

when attempting to predict initiation and propagation

of a particular form of dam age. Toughened resins or

thermoplastics can reduce matrix-dominated damage

but the fibres have the most bearing on impact response

and, over the narrow velocity range under consideration,

the strain rate sensitivity of fibres can be ignored. Post-

impact performance is related to the major damage

mode, therefore a com bination of tension an d compres-

sion residual strength testing is required to characterize

the laminate.

PICS testing is often avoided due to the dilhculty in

providing a large enough gauge section to accomm odate

the damage. This necessitates the use of complex anti-

buckling guides which must support the specimen to

prevent global buckling, but at the same time must not

prevent local instability .

Much research has been performed on simple geo-

metry carbon/epoxy cross-ply lamin ates consisting of

plies at various fibre orientation, due to their importance

in the aerospace industry. The low-velocity impac t

response of random fibre/unidirectional lamina te com-

binations (such as are found in pultrusions) and impacts

on complex geometry are less well documented, and

progress is required in these areas if compo site lamin ates

are to be employed in more structu ral application s.

REFERENCES

Residual flexural strength

Less

work has been

done in this area, but it has been

reported that both flexural mod ulus and strength

decreased with increasing low-velocity im pact energy

for ductile specim ens (glass/epoxy) whilst brittle gra-

phite/epoxy exhibited no losses until complete failure

occurred23. Flexural testing introduces a complex stress

pattern in the specimen; therefore the effect of the

damage on residual strength is less easy to analyse.

Residual fati gue l if e

Jones et a1 79 eported tha t comp ression-comp ression

and tension-com pression are the critical fatigue loading

cases, which would correspond to compression being the

worst-case static loading condition. The maximum resid-

ual comp ressive load divided by the static failure load (s)

typically decreases from 1.0 to 0.6 in the range 1 to lo6

cycles (N), depend ing on the initial dam age size. The rate

of degeneration is at its highest up to N = 100 cycles, and

after lo6 cycles no further degradation occurs; so S = 0.6

may be assum ed to be the fatigue threshold. Therefore it

is believed that fatigue loading is not a good way of

characterizing residual properties.

10

11

12

13

14

15

16

17

CONCLUSIONS

18

19

‘Low-velocity impac t’ refers to impa cts in the range 1 to

Maunsell Structural Plastics Ltd London UK Data Pack

April 1992

Robbins J. Links to a tee. New

Civ il Engineer

August 1992 20-

23

Richmond B.

Bridge construction and

beyond.

Reinforced Pl as-

t ics

October 1993,26-30

Anon. Composite box unveiled.

World Cargo News March 1995

16

Damas P. Genstar Do-It-All.

Containerisafion Int.

April 1995

78- 79

Anon. Looking to a plastic future.

Container M anagement

June

1995 57

Thomas R. More from the floor.

Cargo Syst ems

September

1994 37739

Anon. Envirodek funded.

Cargo Syst ems

March 1995 15

Tooth E.S. Container standardisation: its impact on shipping

and port operations. Dock and Har bour A uthori ty 1990,

70 816), 299-303

Davis R. The protective alternative for cargo containers.

Pro-

duct Fini shing 1988,41 12), 26-27

Anon. Domestic container prospects assessed.

Cargo Systems

In c.: J. ICHCA 1986, 13 6), 81-83

Gooden C.W. Intermodal operators raise efficiency.

Ra i lway

Gazett e Int . 1991. 147 12). 885-888

Munford C. Box fldor”prices hit the roof.

Cargo Syst ems

August 1993 61-63

Anon. Hitting the floor.


July/August

1993 15517

Anon. Floors under the spotlight.

Contai ner M anagement

January 1994 34-35

Anon. Wanted: new floor for new flat.


December 1993 556

Sumerak J.E. and Martin J. Pultruded products-new capabil-

ity on the horizon. In ‘Advanced Composites-Conference Pro-

ceedings’ Dearborn MI 2-4 December 1985 pp. 133-138

Holloway L. Current developments in materials technology and

engineering-pultrusion.

RaI na Review No 3 1989, 2 3) -_

Meyer R.W. ‘Handbook of Pultrusion Technology’ Chapman

and Hall London 1985

29


8/9

Re v i e w o f l o w v e l o c i t y im p a c t p r o p e r t i es : M . 0. W . R i c h a r d s o n a n d M . J . W i s h e a r t

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

Sjoblom, P.O., Hartness, J.T. and Cordell, T.M. On low-velocity

impact testing of composite materials. J.

Compos. M uter.

1988,

22, 30-52

Shivakumar, K.N., Elber, W. and Illg, W. Prediction of low-

velocity impact damage in thin circular laminates.

AZAA J.

1985, 23 3), 442-449

Cantwe ll, W.J. and Morton, J. The impact resistance of compo-

site materials-a review. Composi tes 1991, 22(5), 347-362

Abrate, S. Impact on laminated composite materials.

Appl.

M ech. Rev. 1991,44 4), 155-190

Liu, D. and Malvem, L.E. Matrix cracking in impacted glass/

epoxy plates.

J. Compo s. Mat er.

1987, 21, 594-609

Joshi, S.P. and Sun, C.T. Impact-induced fracture initiation and

detailed dynamic stress field in the vicinity of impact. In ‘Proc.

American Society of Composites 2nd Tech. Conf.‘, DE, 23-25

September 1987, pp. 177 -185

Robinson, P. and Davies, G.A.O . Impactor mass and specimen

geometry effects in low velocity impact of laminated composites.

Znt . J. Zmo act Enz. 1992. 12(2). 189-207

Davies, d.A.0. and Robinson; P. Predicting failu re by debond-

ing/delamination. ‘AGAR D: 74th Structures and Materials

Meeting’, 1992

Joshi, S.P. and Sun, C.T. Impact induced fracture in a laminated

composite.

J. Compos. M uter.

1985, 19, 51-66

Choi, H.Y., Wu, H.-Y.T. and Chang, F.K. A new approach

toward understanding damage mechanisms and mechanics of

laminated composites due to low-velocity impact: Part II-

Analysis.

J. Compos. M uter.

1991, 25, 1012-1038

Jih, C.J. and Sun, C.T. Prediction of delaminatio n in composite

laminates subjected to low velocity impact. J. Compos. M ater.

1993, 27 7), 684-701

Lee, S.-W.R and Sun, C.T. A quasi-static penetration model for

composite laminates.

J. Compos. M uter,

1993, 27(3), 251-271

Cantw ell, W.J. and Morton, J. Geom etrical effects in the low

velocity impact response of CFRP. Compos. Str uct. 1989, 12,

39-59

Wu, H.-Y.T. and Springer, G.S. Measurements of matrix crack-

ing and delamination caused by impact on composite p lates.

J. Compos. M ater.

1988,22, 518-532

Chang,. F.-K., Choi, H.Y . and Jeng, S.-T. Study on impact

damage in laminated composites.

Mech. M ater.

1990,10, 83-95

Choi, H.Y., Downs, R.J. and Chang, F.-K. A new approach

toward understanding damage mechanisms and mechanics of

laminated composites due to low-velocity impac t: Part I

Exneriments.

J. ComDos. M uter.

1991. 25.992-1011

Chbi, H.Y. and Chadg, F.-Y. Impact damage threshold of lami-

nated composites. In ‘Failure Criteria and Analysis in Dynam ic

Response’, AM D Vol. 107, ASME Applied Mechanics Division,

Dallas, TX, November 1990, pp. 31-35

Choi, H.-Y., Wang, H.S. and Chang, F.-K. Effect of laminate

configuration and impactor’s mass on the initial impact damage

of graphite/epoxy composite plates due to line loading impac t.

J. Compos. M uter.

1992, 26(6), 804-827

Chang, F.-K. a nd Chang, K.-Y. A progressive d amage mod el

for laminated composites containing stress concentrations.

J. Comp os. M at er. 1987,21, 834-855

Lagace, P.A. On delamination failures in composite laminates.

In ‘Composite Structures: Testing, Analysis and Design’ (Eds

J.N. Reddy and A.V. Krishna Murty), Narosa Publishing

House, New D elhi, 199 2, pp. 11 1-132

Cui, W. and Wisnom, M.R. A combined stress-based and fracture-

mechanics-based model for predicting delamination in

composites.

Composi t es 1993, 24 6), 467-474

Wang , S.S. Delam ination crack growth in unidirectional fiber-

reinforced com posites under static and cyclic loading. In ‘Com-

posite Materials: Testing and Design (Fifth Conference)‘,

ASTM STP

674 (Ed. S.W. Tsai), American Society for Testing

and Materials, Philadelphia, PA, 197 9, pp. 642-663

Liu, D. Impact-induced delamination-a view of bending stiff-

ness mismatching.

J. Comp os. Mat er. 1988, 22, 674-692

Guild, F.J., Hogg, P.J. and Prichard , J.C. A model for the reduc-

tion in compression strength of continuous fibre composites

after impact damage.

Composi tes 1993, 24 4), 333- 339

Wu, E. and Shyu, K. Response of composite laminates to con-

tact loads and relationship to low-velocity imp act.

J. Compos.

M uter.

1993, 27(15), 1443-1464

Choi, H.T . and Chang, F .-K. A model for predicting dam-

age in graphite/epoxy laminated composites resulting from

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

low-velocity point impact. J. Compos. M uter. 1992, 26(14),

2134-2169

Davies, G.A.O., Z hang, X. and Edlund, A. Predicting damage in

composite aircraft structures due to low velocity impact. ‘Aero-

tech Conference’, Birmingham, UK, January 1 994

Finn, S.R., He, Y.-F. and Springer, G.S. Delamina tions in com-

posite plates under transverse impact loads-experimental

results. Compos. Str uct. 1993, 23, 191-204

Dorey, G . Impact d amage tolerance and assessment in advanced

composite ma terials. Sem inar on Advanced Co mposites, Cran-

field Institute of Technology , C ranfield, UK, 198 6

Dorey, G. Impact damage in composites-development, conse-

quences, an d prevention. In ‘Proc. 6th Int. Conf. on Composite

Materials and 2nd European Conf. on Composite Materials’,

Imperial College, London, 1988, Vol. 3, pp. 3.1-3.26

Dorey, G., Sigety, P., Stellbrink, K. and ‘t Hart, W .G.J. Impact

damage toleran ce of carbon fibre and hybrid laminates. RAE

Technical Report 87 057, Royal Aerospace Establishment,

Farnborough, UK, 1987

Takeda, N., Sierakowski, R.L. and Malvern, L.E. Microscopic

observations of cross sections of impacted composite

laminates.

Compos. Technol . Rev. 1982,4 2), 40-44

Liu, S., Kutlu, Z. and Chang, F.-K. Matrix cracking and delami-

nation in laminated composite beams subjected to a transverse

concentrated line load. J. Compos. M uter. 1993,27(5), 436-470

Garg, AC. D elamination-a damage mode in composite struc-

tures. Eng.

Fract. M ech.

1988, 29(5), 557-584

Finn, S.R. and Springer, G.S . Delamin ations in composite

plates under transverse static or impact loads-a model. Com-

pos. Struct. 1993, 23, 177-190

Dorey, G., Sidey, G.R. an d Hutchings, J. Impact properties of

carbon fibre/Kevlar 49 fibre hybrid composites.

Composi tes

1978,9, 25-32

Clarke, M.P. and Pavier, M.J. Artificial damage techniques for

low velocity impact in carbon fibre composites.

Compos. Str uct.

1993,25, 113-120

Doxsee, L .E., Rubbrecht, P., Li, L., Verpoest, I. and Scholle, M.

Delam ination growth in composite plates subjected to trans-

verse loads.

J. Compos. M at er. 1993, 27 8), 764-781

Razi, H. and Kobayashi, A.S. Delam ination in cross-ply lami-

nated composite subjected to low-velocity impact.

AZAA J.

1993,31 8), 1498-1502

Lee, S. and Zahuta, P. Instrumented impact and static indenta-

tion of composites. J. Compos. M uter. 1991, 25, 204-222

El-Habak, A.M. Effect of impact perforation load on GFRP

composites.

Composi t es 1993,24 4), 341-345

Chaturv edi. S.K. and Sierakowski. R.L. Effects of imnactor size

on impact damage-growth and residual properties in an SMC-

R50 composite.

J. Compos. Mat er. 1985, 19,

100-l 13

Khetan, R.P. and Chang, DC. Surface damage of sheet molding

compound panels subject to a point impact loadin g.

J. Compos.

M ut er. 1983. 17, 182-194

Liu, D., Lee, C.Y. and Lu, X. Repairability of impact induced

damage in SMC composites. J. Compos. Mat er. 1993, 27 13),

1257-1271

Hull, D. ‘An Introduction to Composite Materials’, Cambridge

University Press, Cambridge; 1981

Jang, B.Z., Chen, L.C., Hwang, L.R., Hawkes, J.E. and Zee,

R.H. T he response of fibrous composites to impact loadin g.

Polym. Compos. 1990, 11(3), 144-157

Sela, N. and Ishai, 0. Interlaminar fracture toughness and

toughening of laminated composite materials: a review.

Compo-

sit es

1989, 20(5), 423-435

Moon, F.C . Wave propagation and impact in composite m ateri-

als. In ‘Composite Materials Vol. 7: Structural Design and Ana-

lysis-Part 1’ (Eds L.J. Broutman and R.J. Krock), Academic

Press, London, 1975, pp. 259-332

Caprino , G . Residual strength prediction of impacted C FRP

laminates. J. Compos. Ma ter. 1984, 18, 508-518

Davies, G.A.O., Zhang, X ., Zhov, G. and Watson, S. Numerical

modellina of imoact damage.

ComDosit es 1994.25 5). 342-350

Broutmai, L.J. and R otem, A. Impact strength and’toughness

of fiber composite materials. In ‘Foreign Ob ject Impact D amage

to Composites’,

ASTM STP 568,

American Society for Testing

and Materials, Philadelphia, PA, 1975 , pp. 114-133

Sierakow ski, R.L. and Chaturv edi, S.K. Impact load ing in fila-

mentary structural composites. Shock Vibrat ion Digest 1983,

15 10), 13-31

1130


9/9


2

13

74

15

Caprino, G., Crivelli Visconti, I. and Di Rio, A. Elastic behav-

iour of composite structures under low velocity impact.

Compo-

sir es

1984, 15(3), 23 l-234

Sims, G.D. Understandin g Charpy im pact testing of composite

laminates. In ‘Proc. 6th Int. Conf. on Composite Materials and

2nd European Conf. on Composite Materials’, Imperial College,

London, 1988, Vol. 3, pp. 3.494-3.507

Hayes, S.V. and Adams, D.F. Rate sensitive tensile impact prop-

erties of fully and partially loaded unidirec tional composites.

J.

Tesfing Ed.

1982, 10(2), 61-68

Li, Y.L., Harding, J. and Ruiz, C. Modelling of the impact

response of fibre reinforced composites. Report O U-DE S-

OUEL 1863/90, University of Oxford Department of Engineer-

ing Science, 1990

16

II

78

79

Ruiz, C. and Xia, Y.R. The significance of interfaces in impact

response of laminated composites. In ‘Comp osite M aterial

Technology’, PD Vol. 37, ASME Petroleum Division, New

York, 1991, pp. 161-16 6

Wu, H.-Y.T. and Springer, G.S. Impact induced stresses.

strains, and delaminations in composite plates. J. Compos.

M ater. 1988,22, 5333560

Nettles, A.T. and Hodge, A.J. Compression-after-im pact testing

of thin composite materials. In ‘Proc 23rd Int. SAMPE Tech.

Conf.‘. 21-24 October 1991, pp. 177-183

Jones, R ., Paul, J., Tay, T.E. and Williams, J.F. Assessment of

the effect of impact d amage in composites: some problems and

answers. Compos. Struct. 1988. 10(l). 51 -73

1131

impacto baja velocidad

Documents