0 Stimmen dafür0 Stimmen dagegen

37 Aufrufe10 SeitenNov 25, 2009

© Attribution Non-Commercial (BY-NC)

PDF, TXT oder online auf Scribd lesen

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

37 Aufrufe

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

- Neuromancer
- The E-Myth Revisited: Why Most Small Businesses Don't Work and
- How Not to Be Wrong: The Power of Mathematical Thinking
- Drive: The Surprising Truth About What Motivates Us
- Chaos: Making a New Science
- The Joy of x: A Guided Tour of Math, from One to Infinity
- How to Read a Person Like a Book
- Moonwalking with Einstein: The Art and Science of Remembering Everything
- The Wright Brothers
- The Other Einstein: A Novel
- The 6th Extinction
- The Housekeeper and the Professor: A Novel
- The Power of Discipline: 7 Ways it Can Change Your Life
- The 10X Rule: The Only Difference Between Success and Failure
- A Short History of Nearly Everything
- The Kiss Quotient: A Novel
- The End of Average: How We Succeed in a World That Values Sameness
- Made to Stick: Why Some Ideas Survive and Others Die
- Algorithms to Live By: The Computer Science of Human Decisions
- The Universe in a Nutshell

Sie sind auf Seite 1von 10

composites

M. C. ANDREWS, R. J. Y O U N G

Manchester Materials Science Centre, UMIST/University of Manchester, Grosvenor Street,

Manchester, M1 7HS, UK

Raman spectroscopy has been used to monitor the state of axial stress along fragmented,

high-modulus Kevlar 149 aramid fibres in an epoxy resin matrix by monitoring the peak

position of the strain-sensitive 1610 cm -1 aramid Raman band along individual fragments. It

is shown that the interfacial shear stress along each fragment, derived from the strain-

distribution profiles, is not constant as assumed by conventional fragmentation analysis.

The fragmentation process of as-received Kevlar 149 fibres is compared to that of irradiated

Kevlar 149 fibres exposed to ultraviolet light where the tensile strength and modulus of the

fibres have been reduced. It is found that the derived interfacial shear stress and interracial

shear strength values are higher for those fibres exposed to ultraviolet tight compared with

the as-received fibres. It is also clearly demonstrated that the values of interracial shear

strength calculated at high matrix strains from conventional fragmentation analysis are

considerably lower than the maximum value of interfacial shear stress prior to fibre fracture

that was found to be close to the shear yield stress of the resin matrix. Hence the

determination of the interfacial shear strength following the saturation of the fragmentation

process may give rise to misleading results.

ef Fibre strain material properties which may not give an accurate

em Matrix strain picture of the state of stress in the system. The complex

ep "x Maximum strain along each fragment state of stress due to the loading of the samples can

e~' Failure strain of the fibre often lead to variable, scattered results using con-

Ef Fibre tensile modulus ventional test techniques for the same fibre/matrix

l~ Critical fragment length system [6].

~l~) Mean critical fragment length In the fragmentation test, a single fibre is embedded

If Fragment length in a dumb-bell specimen and aligned parallel to the

r Fibre radius tensile axis. The failure strain of the matrix should be

x Distance along the fibre at least three times the failure strain of the fibre.

~}~"* Maximum stress along each fragment A stress is applied to the sample until failure of the

(y* Fibre tensile strength fibre is observed either optically using polarized light

T Interracial shear stress [3, 7-9] or monitored indirectly by the use of acoustic

9s Interfacial shear strength emission [8]. Continued application of stress results in

the progressive failure of the fibre at several points

along its length until the fragments become so short

1. Introduction that the shear stresses along their lengths can no

Several micromechanical test methods have been de- longer build up enough tensile stresses to cause any

veloped to determine the interfacial properties of further fracture. The interfacial shear strength, rs, can

a composite by measuring the level of interfacial ad- be calculated from

hesion between a reinforcing fibre and a resin matrix.

The methods developed more recently have been

9s - (1)

based upon the deformation of a single fibre partially

or fully embedded in a resin matrix until failure of

either the fibre, matrix or interface occurs. The most where r is the fibre radius, Ic is the critical fragment

common methods include pull-out [1], microbond length and ~* is the tensile strength the fibre at the

[-2, 3], fragmentation [4] and microindentation [5] critical length, Ic. Equation 1 simplifies the analysis of

tests. the fragmentation test in that debonding at the frag-

The level of interracial adhesion is usually charac- ment ends is not accounted for and that the state of

terized by the interfacial shear strength parameter, % stress in each fragment is assumed to be uniform. This

0022-2461 9 i995 Chapman & Hall 5607

is a very simplified approach which can lead to values single value of interfacial shear strength, % calculated

of interracial shear strength being underestimated. following saturation of fibre fracture using conven-

The fibrillar fracture associated with aramid fibres tional analyses [3, 7-9].

make it difficult to determine the fragment length, If,

using conventional techniques [3, 7-9] because the

fractured region may extend axially along the fibre 2. Experimental procedure

making it difficult to determine the point of fracture 2.1. Materials and sample preparation

optically. An alternative method is to measure the A commercial grade of aramid fibre, Kevlar 149, sup-

acoustic emission signals generated by fibre fracture to plied by E.I. Du Pont de Nemours, Wilmington, USA

determine the fragment length [8]. However, both was used in this study. The fibres were ~ 12 gm dia-

optical and acoustic emission methods do not provide meter with an initial fibre modulus, Er, of 145 GPa

any information with respect to the stress in each and a failure strain, e*, of 1.5%. Some of the Kevlar

fragment. 149 fibres were exposed to a low-powered 15 W ultra-

It has been shown that a Raman spectrum can be violet light source in a sealed light box for up to

obtained from a high-modulus Kevlar 149 aramid 5 weeks in order to reduce the tensile strength of the

fibre (Fig. 1) and that the strain-dependent shift of the fibres [13]. The fibres, mounted on paper window

1610 cm-1 aramid Raman band (Fig. 2) can be used cards, were turned over every 24 h. In this way the

to determine the distribution of strain along aramid failure strain of the Kevlar 149 fibres was reduced to

fibres embedded in an epoxy resin matrix [10-12]. less than 1%, so that the failure strain of the epoxy

The technique is extended here to analyse the distribu- resin matrix would be at least three times greater than

tion of fibre strain along a fragmented Kevlar 149 that of the embedded fibres, as is normally required for

aramid fibre during a fragmentation test where the fragmentation tests [-3, 7-9].

fragment length, If, c a n be determined. By monitoring Single short-fibre composite samples were prepared

the distribution of fibre strain at different loading as shown schematically in Fig. 3. Kevlar 149 fibres

levels it is shown that the interracial shear stress, ~, can were cut using ceramic scissors into small sections

be derived at different matrix strains prior to, during, 2-4 mm long. The cut fibres were placed in a two-part

and following fibre fragmentation. The derived values cold-curing epoxy resin matrix consisting of 100 parts

of interracial shear stress can be compared with the by weight LY5052 resin to 38 parts by weight HY5052

hardener supplied by Ciba Geigy. The samples were

cured in a "picture frame" mould for 7 days at room

2000 ~ ' i ' ' i

temperature to avoid any residual thermal stresses in

1329 cm q 1610 cm 1 the composite samples. The resin composite was sub-

1278 cm 1~

sequently cut, shaped and polished to form dumb-bell

1500

shaped specimens ensuring that an aramid fibre was

located in the central region parallel to the tensile axis.

1517 cm ~

< 1000 The tensile behaviour of the resin matrix was also

determined. Values of 3.1 GPa, 70 MPa and 4.1%

1180 c m -1

were measured for the matrix tensile modulus, yield

500 stress and failure strain, respectively. The compressive

yield stress of the resin was found to be about

105 MPa which means that the shear yield stress of

1200 1300 1400 1500 1600 the matrix, Zy, is about 43 MPa [14].

Wave number, Av (cm ~)

1700 cm -1. 2.2. Raman spectroscopy

The single-fibre composite samples were mounted in-

dividually on a Miniature Materials straining rig,

1611 I I I I I i i (Polymer Laboratories Ltd, UK). The strain in the

U matrix was measured using a strain gauge of gauge

E 1610

v

O factor 2.09 bonded to the surface of the test region

> 1609 using a cold-curing slow-setting epoxy resin adhesive,

e =_5.1cm_1/%

ff 1608 as shown schematically in Fig. 3.

O

"~ 1607

O

1606

Resistance

2 1605 strain gauge

c

1604 Fibre /J

1603

' 0:0 ' 0:2 " 0'.4 0'.6 ' 0'.8 ' 1'.0 ' 1'.2 " 1.4 Applied

stress, ~m

Fibre strain, el( % )

band upon fibre strain for a Kevlar 149 fibre. Figure 3 Schematic diagram of a single-fibre model composite.

5608

The samples were placed on the stage of an Olym- 1.6

pus BH-2 optical microscope connected via a set of 1.4 +

+

collection optics to a Spex 1000M single mono- 1.2

+ ++ , ,+** I

+ + +++

chromator. Raman spectra were obtained during de- oooo

1.0

formation of the fibres in air and at regular intervals =- V v

along the aramid fibres in the model composite sam- 0.8 vv

9v v ' ' v v v v v

- - - VY

E

ples, using the 632.8 nm red line of a 15 m W he- 0.6

lium-neon laser focused to aia ~ 5 gm spot on the 0.4 I 9

Ls m I i I []

surface of the fibre. A highly sensitive Wright Instru- 0.2

ments charge-coupled device (CCD) camera cooled 9 . 9 |

0.0 o o 9 o " g " o

with liquid nitrogen was used to collect the Raman

-0.2

spectra using an exposure time of less than 10 s. For 0 200 400 600 80(

each sample, the peak position of the strain sensitive (a) Distance along fibre, x ( #m )

1610 cm -1 Raman band [15, 16] was used to map the

fibre strain distribution profiles along the length of the

50

embedded fibres at matrix strains, era, from 0%-2.0% em = 10.8% r , i

l

using small 10 gm steps close to the fibre ends.

40

era=0 1 ~ o

3. Results and discussion 13_ 30 \ \/era = 1.0%

3.1. Stress transfer in a Kevlar 149

20 em=0" l ~ em = 1"2~176

single-fibre model composite if)

(/3

The distribution of strain along the left-hand end of

a single Kevlar 149 fibre in the epoxy resin matrix 10

calculated from the point-to-point variation of the

1610 cm-1 aramid Raman band is shown in Fig. 4a. 0

In the'unstrained case, em= 0%, there is virtually no u L n

200 400 600 800

strain in the fibre. As the matrix strain is increased the

s t r a i n in the fibre increases from the fibre end up to (b) Distance along fibre, x ( lurn )

a plateau strain value, along the middle of the fibre, Figure 4 (a) Variation of fibre strain and (b) derived variation of

approximately equal to the matrix strain. This behav- interracial shear stress, z, with distance along the left-hand end of

iour is similar to that predicted by shear-lag theory a Kevlar 149 fibre in a single-fibre model composite at different

[17]. However, unlike the behaviour predicted by levels of matrix strain. (a) era; (O) 0%, (ll) 0.4%, (A) 0.6%,

(Y) 0.8%, (~) 1.0%, ( + ) 1.2%.

shear-lag theory, the strain at the fibre end, x = 0 }am,

is not equal to zero. This type of behaviour has also

been reported for other aramid/epoxy systems

where em= 0.8% and is close to the shear yield stress

[10-12, 18] and has been attributed largely to the

of the resin matrix. This behaviour has been repeated-

geometry of the cut fibre end [19]. It is also shown

ly observed [10, 14, 18] for several aramid fibre/epoxy

that the solid lines in Fig. 4a, which are a fit of the

model composite systems and indicates that the max-

experimental data to a logistic sigmoid function

imum interracial shear stress is highly dependent upon

[10-12, 18], tend to smooth the experimental data

the mechanical properties of the epoxy resin matrix.

particularly along the central region of the fibre. The

Localized matrix yielding at the fibre/matrix interface

scatter in the data could be attributed to either inac-

causes the maximum value of interfacial shear stress to

curacies in the curve-fitting procedure used to deter-

move along the fibre and can also initiate interfacial

mine the central peak position of the 1610cm - t

debonding close to the fibre end. This results in a re-

Raman band or may represent locally stressed regions

duction in the values of interfacial shear stress at the

along the length of the fibre where fibre fracture may

fibre end, as observed in Fig. 4b at 1.0% and 1.2%

be initiated.

matrix strain.

The interfacial shear stress, % can be derived using

a force balance equilibrium [20] given by

"C = Ef" (2) Kevlar 149 fibres

Figs 5 and 6 show the distribution of fibre strain along

where d e d d x is the differential of the variation of fibre the entire length of a Kevlar 149 fibre in a model

strain with distance aIong the fibre calculated from the composite at 0.5% and 1.0% matrix strain. At the low

solid line fitted to the data in Fig. 4a. The variation of strain level, e~ = 0.5%, the behaviour is similar to that

interfacial shear stress with distance, x, along the Kev- predicted by shear-lag theory as was shown in Fig. 4a.

lar 149 fibre in Figl 4a is shown in Fig. 4b. At matrix It is shown, however, that the fibre strain fluctuates

strains up to 0.8% the interracial shear stress is a max- along the middle of the fibre. As discussed earlier, this

imum at the fibre end and decreases to zero at a dis- may be due to the error in determining the peak

tance, x, along the fibre. The interfacial shear stress at position of the 1610 cm -1 aramid Raman band or

the fibre end reaches a maximum value of 42 M P a may represent local variations in fibre strain along the

5609

I , I .,,. q , I I I ,

[

1.5 1.5

A

o~

1.0 1.0

=-

X

0.5 0,5

..Q

LE L~

0.0 0.0

i L I i ' I ' i

0 500 1 0 0 0 1 5 0 0 2000 2500 6 5()0 113'00 15'00 20'00 ' 25'00 '

Distance along fibre, x(ptm ) (a) Distance along fibre, x ( }am )

fibre in a single-fibre model composite at (0) 0% and (in) 0.5% 30 I I I L ~ ~ [

strain.

20

1 ] I t I I I

10

n

1.5 0

O oq

w

-10

1.0

e-

-20 -

0.5 -30

LI_

; 500 10'00 15'00 20~00 2500

(b) Distance along fibre, x ( ~m )

0.0 ~-'- .... --~ "- -'~ , o '~o-O

Figure 7 (a) Variation of fibre strain along an as-received Kevlar

149 fibre in a single-fibre model composite at (O)0% and

(} 5;0 1dO0 ' 1500 ' 20'00 25'00 (x) 1.5% strain. (b) Derivedvariation of interracial shear stress, r,

Distance along fibre, x ( grn ) along an as-receivedKevlar 149 fibre in a single-fibremodel com-

posite at 1.5% strain.

Figure 6 Variation of fibre strain along an as-receivedKevlar 149

fibre in a single-fibremodel compositeat (0) 0% and (~) 1.0%

strain.

located close to the fibre ends and at points of fracture

along the fibre. The dotted lines in Figs 7b and 8b

fibre. At 1.0% matrix strain the fibre strain distribu- indicate the average values of maximum interracial

tion in Fig. 6 is trapezoidal due to yielding of the resin shear stress at the fibre and fragment ends.

matrix close to the fibre ends [4, 21]. It is apparent that the interracial shear stress along

When the matrix strain is increased to 1.5% strain, each fragment is not constant as assumed by conven-

as shown in Fig. 7a, the failure strain of the fibre is tional fragmentation analysis [4] and presumed by

Equation 1. Interfacial failure and localized matrix

exceeded and fibre fracture occurs with the fibre

yielding at the interface can lead to non-linear distri-

breaking into two fragments. The fibre strain increases

butions of fibre strain along the fibre fragments that

from the fibre ends in each of the fragments to a value

are not expected by conventional fragmentation anal-

of 1.4%, i.e. less than the matrix strain, at which point

ysis which predicts that the distribution of fibre

the fibre strain decreases. Increasing the matrix strain

strain along each fragment should be approximately

to 2.0% results in one of the fragments breaking into

triangular.

two smaller sections, shown in Fig. 8a. It is also shown

that the maximum strain in each fragment has de-

creased even though the matrix strain has increased.

The solid lines in Figs 5, 6, 7a and 8a have been 3.3. E f f e c t o f u l t r a v i o l e t l i g h t u p o n t h e

fitted using a cubic spline function [22]. This enables t e n s i l e s t r e n g t h o f K e v l a r 149 a r a m i d

the variation of interracial shear stress with distance, fibres

x, along the fibre to be derived accurately, using Equa- The strain to failure, e~, of aramid fibres is relatively

tion 2. This is shown in Figs 7b and 8b for matrix high compared with other high-performance fibres.

strain levels of 1.5% and 2.0% where fibre fragmenta- The strain to failure of most aramid fibres is in excess

tion has occurred. When fibre fragmentation occurs, of 2.5%. Therefore, most aramid fibres are unsuitable

the interracial shear stress distributions in Figs 7b and for studying fibre fragmentation in the present resin

8b become more complex than at low strains (Fig. 4b) matrix where the failure strain is only 3-4%. The

with maximum values of interfacial shear stress failure strain for the high-modulus fibre, Kevlar 149, is

5610

r I I ~ f r I i t

3.5

1.5 3.0

A

*~ 2.5 7- T 1 0 *ff

1.0 0 8 .~

=" 2.0-

Q

1.5-

0.5 "~ 0.4

~3

LE 1.o-

0.2

0.0 0.5 .................................... 0.0

0 7 14 21 28 35

Exposure time ( days )

b 500 1600 1570 20'00'25'00

(a) Distance along fibre, x ( :am ) Figure 9 Effectof ultraviolet light upon the (11) tensiIestrength and

(9 strain to failure of Kevlar 149 fibres.

n

n 160

150

o9

03

140

al fibre modulus

-o 130

0

-20 - E 120

.13

LT. 110

-30

0 500 1000 1500 2000 2500 100

(b) Distance along fibre, x ( Fm )

90 ' t . . . . . . t . . . . . . I . . . . . . ~ . . . . . . f . . . . . .

0 7 14 21 28 35

Figure 8 (a) Variation of fibre strain along an as-received Kevlar

Exposure time ( days )

149 fibre in a single-fibremodel composite at (Q) 0% and (,) 2.0%

strain. (b) Derivedvariation of interracial shear stress, x, along an Figure 10 Effectof ultraviolet light upon the initial tensile modulus

as-received Kevlar 149 fibre in a single-fibre model composite at and the tensile modulus prior to failure of the Kevlar 149 fibres.

2.0% strain.

model composite to be deformed in excess of the

failure strain of the fibre.

The effect of ultraviolet radiation upon the tensile

properties of aramid materials has been described

previously [13, 23]. The aramid polymer degrades due

to chain scission of the amide linkage in the polymer

backbone [24].

The tensile properties of Kevlar 149 fibres exposed

to a low-powered ultraviolet light source have been

investigated in this present study, Fig. 9 shows the

reduction in the tensile strength, cy~~, and the failure

strain, e~, for Kevlar 149 fibres exposed to ultraviolet Figure ll Scanning electron micrograph showing the wedge-

light. A 50% reduction in both the tensile strength and shaped spiral defects along the length of an unexposed, as-received

strain to failure was observed following exposure of Kevlar 149 fibre.

the Kevlar 149 fibres to the ultraviolet source for

a period of 5 weeks. This was accompanied by a reduc-

tion in the initial fibre modulus and also the fibre fibres more susceptible to photodegradation com-

modulus prior to tensile failure, as shown in Fig. 10. pared with other aramid fibres E13] that do not pos-

The tensile modulus prior to failure was determined sess such structural defects. The extensive damage

from the tensile strength and strain to failure given by caused by photodegradation results in a reduction of

the solid lines in Fig. 9. the tensile properties of the aramid fibres. The irra-

It is thought that wedge-shaped helical defects diated fibres fail in tension due to premature fibrillar

along the length of the unexposed as-received Kevlar fracture along the helical defect regions. The fibrillar

149 fibres, shown in Fig. 11, allow the ultraviolet light regions between the wedge-shaped defects are prone

to penetrate into the centre of the fibre and make the to extensive damage. When this region is loaded in

5611

tension, fibrillar fracture can occur that allows shear- ..... I , , t t r J I , i I

1.5

thought to account for the reduction in the initial

tensile modulus [13]. o

tion behaviour of irradiated Kevlar 149 fibres in or

model composites has been investigated and is de- '~

veloped next. ~ 0.5

..Q

1.1-

3.4. Fragmentation of Kevlar 149 fibres o.0 -- "br'b~-',,'~ " " -- ~ ~ - ' ~ B - - vo'Ir -,,,',mr 9

If fibre fracture occurs at low strain levels (less than 0 1000 2000 3000 4000

1%) the influence of localized matrix yielding and Distance along fibre, x ( grn )

interracial debonding [14, 18] upon the fragmentation

process are reduced. This can be achieved by analys- Figure 13 Variation of fibre strain along a Kevlar 149 fibre, ex-

ing the fragmentation process of the Kevlar 149 fibres posed to ultraviolet light, in a single-fibre model composite at

exposed to ultraviolet light for 35 days where the (O) 0% and (O) 1.0% strain.

strain to failure of the fibres has been reduced to less

than 1.0%, as shown in Fig. 9. Figs 12, 13, 14a and

15a show the variation of fibre strain along an irra- I i I i I , I I I

different levels of matrix strain up to 2.0%. The distri-

bution of fibre strain at 0.5% matrix strain is similar o~

to that shown in Fig. 5 for the as-received fibre model ~ 1.0.

composite. At 1.0% matrix strain (Fig. 13) initial fibre ="

failure has occurred 1100 gm from the left-hand end of .~

the fibre. Because the fibre strain at x = 1100 ~tm does .~ 0.5.

not decrease to zero, it may be assumed that the crack -Q

LE

has not propagated totally across the diameter of the

fibre. As the matrix strain is increased to 1.5% 0.0-

(Fig. 14a) it can be seen that the fibre strain at

x = ll00~tm has decreased to 0,25% indicating ; lObO 20'00 3doo 40'00

propagation of the crack across the fibre. It is also (a) Distance along fibre, x ( gm )

shown that further breaks have occurred randomly

along the length of the fibre. This process continued

until at 2.0% matrix strain (Fig. 15a) the fibre frag- 40

mented into a number of small fragments of length, If,

in which the distribution of strain is approximately 20

triangular. The m a x i m u m strain in each fragment,

e~ ax, varies between 0.6% and 1.4% strain. It is evi- o_

dent that the m a x i m u m strain at different points along "s 0

the fibre is much lower than the strain in the matrix. It r

-20

I L I i I I I i r

- 40

1.5

0 1000 2000 3000 4000

(b) Distance along fibre, x ( g m )

cff 1.0

Figure 14 (a) Variation of fibre strain along a Kevlar 149 fibre,

exposed to ultraviolet light, in a single-fibre model composite at

0.5 (Q) 0% and (x) 1.5% strain. (b) Derived variation of interracial

shear stress, z, along a Kevlar 149 fibre, exposed to ultraviolet light,

ft.

in a single-fibremodel composite at 1.5% strain.

0.0-

i i

0 lOOO 20'oo 30'oo 40.oo therefore follows that the fibre in Fig, 15a is offering

Distance along fibre, x ( ~m ) very little reinforcement.

The cubic splines fitted to the curves in Figs 14a and

Figure 12 Variation of fibre strain along a Kevlar 149 fibre, ex- 15a have enabled the distribution of interracial shear

posed to ultraviolet light, in a single-fibre model composite at stress to be derived along the length of the fragmented

(0) 0% and (I) 0.5% strain. fibre using Equation 2. The interracial shear stress in

5612

I I I ( I ) I

[3, 7-9] rely upon the precise determination of the

1.5

critical length fragment, lc, in order to calculate the

interfacial shear strength, ~cs, using Equation 1 [4].

o~ This assumes that the shear stress along the interface

~T 1.0 is constant. It is shown in Fig. 15a that while the

d distribution of fibre strain along each fragment is

approximately triangular, the interfacial shear stress

0.5 in Fig. 15b, derived from the fitted curve in Fig. 15a, is

.D

u.

by no means constant along each fragment. If the

interracial shear stress along each fragment is assumed

0.0 to be constant, then the change in fibre strain along

each fragment should be approximately linear thus

'

1000 '

2000 30 ; 0 40'00 forming a triangular fibre strain distribution along the

(a) Distance along fibre, x ( tam ) length of the fibre. This is demonstrated in Fig. 16a

where linear regression curves have been fitted to the

40 strain distributions along each fragment in Fig. 15a.

The variation of interracial shear stress derived from

the straight lines in Fig. 16a is shown in Fig. 16b. The

2O

uniform values of interfacial shear stress are approx-

r imately equal to +10 MPa and are considerably

0 lower than the values derived in Fig. t5b from the

O9

cubic spline curve fitted with greater accuracy to the

O9 same fibre strain distribution data.

-20 If the critical fragment length, Ic, can be determined

it is possible to determine a value of interfacial shear

- 40

strength, ~s, using Equation 1. The critical fragment

length, Ic, is often underestimated if it is assumed to be

1000 2000 3000 4000 equal to the fragment length, If. Towards the end of

(b) Distance along fibre, x ( tam ) the fragmentation process, fragments in the range

Ic < If < 21o will continue to break to produce frag-

Figure 15 (a) Variation of fibre strain along a Kevlar 149 fibre,

ments in the range Io/2 < If < l~. Therefore, a uniform

exposed to ultraviolet light, in a single-fibre model composite at

(0) 0% and (*) 2.0% strain. (b) Derived variation of interfacial

fragment length distribution is often assumed with

shear stress, ~, along a Kevlar 149 fibre, exposed to ultraviolet light, a mean value ( l o ) = 0.75lo [3, 7-9]. The interracial

in a single-fibre model composite at 2.0% strain. shear strength is now given by

3 r~

~s - (3)

each fragment has a maximum value at the fragment 4 (Ir

ends. Along the middle of the fragments smaller peaks

A better estimate of the mean critical fragment length

in interracial shear stress can be observed that may be

a result of partial fibrillar fracture along the fragment. can be obtained by using a Monte-Carlo simulation of

The dotted lines in Figs 14b and 15b represent the a Poisson/Weibull probability model [25, 26] to rep-

resent the random occurrence of flaws along a fibre

mean value of maximum interracial shear stress

with a linear build up of the fibre tensile stress away

located at the fragment ends and is approximately

equal to ___22MPa. This is higher than the mean from the fibre breaks.

value of maximum interfacial shear stress for the as- Another problem associated with the conventional

analysis is the determination of the tensile strength of

received Kevlar 149 fibre system that was approxim-

the fibre at the critical fragment length because the

ately equal to + 14 MPa as shown in Figs 7b and 8b.

This implies that the irradiated fibres have better fibre tensile strength usually increases as the fragment

adhesion with the epoxy resin matrix than the as- length decreases [27]. The value of fibre tensile

received fibres. This may be possible if the free radicals strength is often calculated by measuring the tensile

generated during chain scission react to form oxidized strength of the fibres with different gauge lengths

chain ends that improve the chemical bonding at the (2-200 mm) and extrapolating back to find the tensile

fibre/matrix interface. The increased level of adhesion strength of the fibre at the critical fibre fragment

and reduction in tensile strength both lead to shorter length, l~. Conventional analysis also assumes that the

fibre fragments in the case of the irradiated Kevlar 149 fibre strain at the point of fracture is equal to zero.

fibres. This can be seen in Table I where the fragment This does not appear to be the case in Fig. 15a where

lengths for the as-received fibres are longer than those the strain at the fragment ends is between 0% and

for the irradiated fibres. 0.2%.

If the maximum stress in each fragment can be

determined then the interfacial shear strength can be

calculated by modifying Equation 1 to give

3.5. Conventional fragmentation analysis

Conventional fragmentation procedures using either r O~ lax

9s - (4)

acoustic emission [8] or polarized light techniques If

5613

T A B L E I Values of interracial shear strength for the as-received and irradiated Kevlar 149 fibres calculated from the individual fragment

lengths, shown in Figs 7a, 8a and 15a, using Equation 4

Fibre Matrix strain Fragment length Maximum strain Maximum stress Interracial

em (%) If (gm) e~"x (%) o~ "~ (GPa) shear strength

z~ (MPa)

em= 1.5% 1250 1.5 2.18 10.7

Fig. 8a 1300 1.0 1.45 6.7

em = 2.0% 810 0.7 1.02 7.6

470 0.5 0.73 9.3

Mean 8.6 (1.4)

Irradiated Kevlar 149 Fig. 15a 400 0.5 0.73 11.0

em = 2.0% 700 0.9 1.31 11.2

1050 1.3 1.89 10.8

750 0.9 1.31 10.5

750 0.9 1.31 10.5

550 1.2 1.74 18.9

Mean 12.2 (3.1)

and 15a not only can the fragment length, If, be meas-

1.5

ured but also the maximum strain in each fragment,

ep ax, is known from which the maximum stress, a~ ax,

can be calculated. The interfacial shear strength, % 1.0

can be calculated, using Equation 4, for each fragment

shown in Figs 7a, 8a and 15a where there is well-

defined saturated fragmentation. These values are 0.5

given in Table I.

LZ.

The mean interracial shear strength, %, was cal-

culated using Equation 4 for both the as-received 0.0 - i r O , , I q r - - m r ~ " l l l w r,~IIFaD~ IIpm~lD- - ~ 9

composite samples and the values are given in Table I. ; 10'00 20'oo 30'oo 40'oo

The mean value of interracial shear strength of (a ) Distance along fibre, x ( grn )

12.2 M P a for the irradiated Kevlar 149 fibre is, as

expected, close to the maximum values of interfacial 30 i , i , ' i , i , i ,

angular fibre strain distribution, shown in Fig. 16a, 20

where the interfacial shear stress is assumed to be

constant. It is also interesting to note that the mean 10

n

interracial shear strength of t2.2 M P a calculated for

the irradiated Kevlar I49 sample is higher than the v

0

mean value of zs of 8.4 M P a calculated for the as- U)

received Kevlar 149 fibre samples. This was also found -10

to be the case for the values of interracial shear stress

derived from the Raman fibre strain distributions in- -20

dicating that the irradiated fibres have better adhesion

than the as-received fibres. -30 i , i , i , i , i '

The mean values of interfacial shear stress, z, and

(b) Distance a l o n g fibre, x ( ~m )

interracial shear strength, % appear to be considerably

lower than the values of maximum interracial shear

Figure 16 (a) Variation of fibre strain along the Kevlar 149 fibre

stress, z .... shown in Fig. 4b, calculated at low strain exposed to ultraviolet light in Fig. 15a with linear regression lines

levels, em ~< 1.2% for an as-received Kevlar 149 model fitted to the strain data along each fragment, era; (0) 0%, (*) 2.0%.

epoxy composite. This is clearly shown in Fig. 17 (b) Variation of interracial shear stress, ~, derived from the linear

which gives the values of "~maxfrom Fig. 4b for the regression lines in (a) along the Kevlar 149 fibre exposed to ultra-

Kevlar 149 model composite together with the data violet light.

calculated from fragmentation analysis. The circled

data at 1.5% and 2.0% matrix shown in Fig. 17

indicate the mean values of maximum interfacial shear interracial shear strength values shown in Table I cal-

stress at the fragment ends indicated by the dotted culated from Equation 4 using the conventional frag-

lines in Figs 7b and 8b and Figs 14b and 15b for the mentation analysis. The interfacial shear strength cal-

as-received and irradiated Kevlar 149 fibres, respec- culated from the fragmentation data was determined

tively. The dotted lines in Fig. 17 indicate the mean following fibre fracture which usually occurs after the

5614

_ _ 1 i ~ i

Matrix shear yield stress, ~y

I , I

apparent from the values of interracial shear strength,

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

z~ calculated from the saturated fragmentation data

which are calculated only at high strain levels.

The use of Raman spectroscopy to study the frac-

j 30 adia~149 ture of aramid fibres in a resin matrix has the advant-

age that the state of axial stress or strain can be clearly

-- 20

defined along the fragmented regions. This should be

compared with conventional fragmentation tech-

.fi ted K~.vJ.a.r__?49 .~ce.,. .!_2.:2. niques [3, 7-9] that do not give any direct measure of

the state of stress in the fragments and in the case of

As-received Kevlar 149 aramids even the lengths of the fragments are difficult

0 to measure. Using polarized light it is, however, pos-

o.o o'.a 2:0 sible to observe fragmented regions in a model com-

Matrix strain, em ( % ) posite and then map the distribution of strain along the

fragments using Raman spectroscopy. This is demon-

Figure 17 Dependence of maximum interfacial shear stress, ~.... strated in Fig. 18 which shows the fibre strain distribu-

upon matrix strain for a Kevlar 149 fibre in a single-fibre model

composite. (...) Values of interracial shear strength, % calculated tion along the fragmented, irradiated Kevlar 149 fibre,

from the fragmentation data using Equation 4. shown in Fig. 15a, along with the birefringence pattern

associated with fibre fracture in the same model com-

posite viewed using polarized light. The problems

maximum value of interracial shear stress has been associated with conventional fragmentation analysis

attained. The maximum value of interfacial shear discussed earlier, however, still remain.

stress occurs prior to the onset of matrix yielding at

the fibre/matrix interface [14, 18], i.e. when

em ~ 0.8%, after which point the value of ~max de- 4. Conclusions

creases, as shown in Fig. 17. This indicates that the It has been shown that Raman spectroscopy can be

maximum interfacial shear stress is dependent upon used to define clearly the axial strain-distribution pro-

the shear yield stress of the resin [14]. This is not files along fragmented aramid fibres from which the

Figure 18 Variation of fibre strain along the Kevlar 149 fibre in Fig. 15a, exposed to ultraviolet light, in a single-fibre model composite at

(0) 0% and (*) 2.0% strain and inset an optical micrograph of the same sample viewed under cross-polarized light.

5615

interfacial shear stress can be derived accurately. It 2. B. MILLER, P. MUR] and L. REBENFELD, Compos. Sci.

was found that the values of interfacial shear stress, "c, Technol. 28 (1987) 17.

3. P.J. HERRERA-FRANCO and L. T. DRZAL, Composites 23

and interfacial shear strength, % were higher for the

(1992) 2.

Kevlar 149 fibres exposed to ultraviolet light and may 4. A, KELLY and W. R. TYSON, J. Mech. Phys. Solids 13 (1965)

be a result of additional interfacial interactions be- 329.

tween oxidized chain ends, formed by the reaction of 5. J.E. MANDELL, D. H. GRANDE, T. H. TSIANG and F. J.

free radicals generated during chain scission, and the MdGARRY, in "Composite Materials Testing and Design"

epoxy resin matrix. (Seventh Conference), ASTM STP 893, edited by J. M. Whit-

ney (American Society for Testing and Materials, Philadel-

Using the Raman technique it is shown that each phia, PA, 1986) p. 87.

fragment length can be measured from the strain dis- 6. M. J. PITKETHLY, J. P. FAVRE, U. GAUR, J.

tribution profiles that would otherwise be difficult JAKUBOWSKI, S. F. MUDRICH, D. L. CALDWELL, L.

using conventional polarized light methods due to the T. DRZAL, M. NARDIN, H. D. WAGNER, L. DI LAND-

RO, A. HAMPE, J. P. ARMISTEAD, M. DESAEGER and

fibrillar fracture associated with aramid fibres.

I. VERPOEST, Compos. Sci. Technol. 48 (1993) 205.

Determination of the interfacial properties of 7. H . D . WAGNER, H, E. GALLIS and E. WIESEL, J. Mater.

aramid fibres in an epoxy resin matrix using conven- Sci. 28 (1993) 2238.

tional fragmentation methods and analyses is usually 8. A . N . NETRAVALI, Z-F LI, W. SACHSE and H. F. WU,

difficult, because the fragment length and tensile ibid. 28 (1991) 6631.

9. J. KALANTAR and L. T. DRZAL, ibid. 25 (1990) 4194.

strength of the fragment cannot be easily determined.

10. M.C. ANDREWS and R. J. YOUNG, Mater. Sci. Eng. A184

It is also apparent that the assumption of constant (1994) 197.

interracial shear stress is too simplified in the conven- 11. M.C. ANDREWS and R. J. YOUNG, J. Raman Speetrosc. 24

tional analysis and may not be appropriate. This is (1993) 539.

because the values of interfacial shear strength cal- 12. M . C . ANDREWS, R. J. DAY, A. K. PATRIKIS and R. J.

culated at high matrix strains following fibre fracture YOUNG, Composites 25 (1994) 745.

13. M. G. DOBB, R. M. ROBSON and A. H. ROBERTS, J.

are considerably lower than the maximum value of Mater. Sci. 28 (1993) 785.

interfacial shear stress calculated prior to fragmenta- 14. M.C. ANDREWS, R. J. Y O U N G and R. J. DAY, in "Devel-

tion. This was found to be close to the shear yield opments in the Science and Technology of Composite Mater-

stress of the epoxy resin matrix. This implies that ials", edited by A. R. Bunsell, A. Kelly and A. Massiah (Wood-

interfacial failure is controlled by yielding of the epoxy head, Cambridge, 1993) p. 133.

15. S. VAN DER ZWAAG, M. G. NORTHOLT, R. J. YOUNG,

resin matrix at the fibre/matrix interface and occurs I. M. ROBINSON, C. GALIOTIS and D. N. BATCHEL-

well before the fragmentation process saturates. Hence DER, Polym. Commun. 28 (1987) 276.

great care must be exercised in the determination of 16. C. CHANG and S. L. HSU, Macromolecules 23 (1990) 1484.

interracial parameters using conventional fragmenta- 17. II. L. COX, J. Appl. Phys. 3 (1952) 72.

tion analysis in the "saturated" state, because inter- 18. M. C. ANDREWS, R. J. YOUNG and J. MAHY, Compos.

Interfaces, 2 (1994) 433.

facial failure may have occurred at significantly lower 19. C . F . FAN and S. L. HSU, Macromolecules 22 (1989) 1474.

levels of matrix strain. 20. A. KELLY and N. H. M acMILLAN, "Strong Solids", 3rd Edn

(Clarendon Press, Oxford, 1986).

21. M. R. P I G G O T T , "Load Bearing Fibre Composites" (Per-

Acknowledgements gamon, Oxford, 1980) p. 83.

22. C.H. REINSCH, Numer. Math. 10 (1967) 177.

This work was supported by a Research Grant from 23. S. C. POWELL, R. L. KIEFER, P. L. PATE and R. A.

the Science and Engineering Research Council and the ORWOLL, Polym. Prep. 32 (1991) 122.

Ministry of Defence. The authors thank E.I. Du Pont 24. D . J . CARLSSON, L. H. GAN and D. M. WILES, J. Polym.

de Nemours for supplying the aramid fibres used in Sci. A. Polym. Chem. 16 (1978) 2353.

25. W . F . KNOEE, J. Mater. Sci. 28 (1993) 931.

this study. R. J. Y. thanks the Royal Society for sup-

26. R.B. HENSTENBERG and S. L. PHOENIX, Polym Compos.

port in the form of the Wolfson Research Professor- 10 (1989) 385.

ship in Materials Science. 27. S.A. FAWAZ, A. N. PALAZOTTO and C. S. WANG, Poly-

mer 33 (1992) 100.

References

1. L.J. BROUTMANN, "Interfacesin Composites", ASTM STP

452 (American Society for Testing and Materials, Philadel- Received 22 December 1994

phia, PA, 1969) p. 27. and accepted 15 March 1995

5616

- 19841 (6)Hochgeladen vonFMATEO
- BME311-reading assignment-2 stress strain.pdfHochgeladen vongouravgoelshinu
- 81.IJMPERDFEB201881Hochgeladen vonTJPRC Publications
- Mechanical properties of Aluminum-Copper(p) composite metallic materials.pdfHochgeladen vonBalaChandar
- C-27 Piping Flexibility AnalysisHochgeladen vonChuy Lee
- IJAIEM-2014-10-01-92Hochgeladen vonAnonymous vQrJlEN
- Optimum Design and Analysis of Electric Kart ChassisHochgeladen vonInternational Journal of Innovative Science and Research Technology
- 50_Ministerial Roll - Madras DioceseHochgeladen vonIsaac Samuel
- vitezHochgeladen vonDattatraya Parle
- Sub. CompoziteHochgeladen vonCristina-Alexandra Crăciun
- Unconfined Compression Strength - Set 1Hochgeladen vonahm531
- Project 12Hochgeladen vonPrashit Nayak
- Improved Processing of Carbon Nanotube-magnesium Alloy CompositesHochgeladen vonMárcio Martins
- Measurement of Meat TextureHochgeladen vonsitinurhaniza
- armor bhat.pdfHochgeladen vondharm81
- RDS 1 (T-2403)Hochgeladen vonSHANMUGAM V
- RISK_2_1_011001-FinalHochgeladen vonMohamed AlaaElDin
- Welded Joint Definition of Welded Joint in the Free Online EncyclopediaHochgeladen vonchowhk
- 802694_ch9Hochgeladen vonmvjayaram007
- Dissertation Schoettler 3412262[1]Hochgeladen vonJamesFarrugia
- mAKALEHochgeladen vonOktay
- 6 2 Shearing StressesHochgeladen vonMohd Danial H
- Fallas Por Fatiga Criterios y ProblemasHochgeladen vonTello Est
- All ReportsHochgeladen vonleonardomayer4350
- Lecture 2-Strength Analysis of Beams Accdg to AciHochgeladen vonMarcelo Abrera
- Block ShearHochgeladen vonSuhas Natu
- Retroffitng RC Steel-MRF WCEE2012 2824Hochgeladen vongorgika papand
- Asme Case 2120-1Hochgeladen vonDinh Nguyen Gia
- MENG-4AHochgeladen vonEd Lewis
- fulltext_011.pdfHochgeladen vonkhamis

- t6Hochgeladen vonapi-3733260
- Kama Sutra - Sex Positions - EnglishHochgeladen vonsteriandediu
- wangHochgeladen vonapi-3733260
- zeronianHochgeladen vonapi-3733260
- VSUBRAMHochgeladen vonapi-3733260
- YESIMHochgeladen vonapi-3733260
- yeoHochgeladen vonapi-3733260
- yearsHochgeladen vonapi-3733260
- yannasHochgeladen vonapi-3733260
- yangHochgeladen vonapi-3733260
- wortmannHochgeladen vonapi-3733260
- WOOLHochgeladen vonapi-3733260
- whiteHochgeladen vonapi-3733260
- WETTONHochgeladen vonapi-3733260
- weiHochgeladen vonapi-3733260
- warwickerHochgeladen vonapi-3733260
- vinisHochgeladen vonapi-3733260
- vijayanHochgeladen vonapi-3733260
- vershHochgeladen vonapi-3733260
- ULLMANHochgeladen vonapi-3733260
- ugbolueHochgeladen vonapi-3733260
- tsurHochgeladen vonapi-3733260
- treloarHochgeladen vonapi-3733260
- tereneHochgeladen vonapi-3733260
- taborHochgeladen vonapi-3733260
- stressHochgeladen vonapi-3733260
- strHochgeladen vonapi-3733260
- stearneHochgeladen vonapi-3733260

- 14 Grams to Moles to MoleculesHochgeladen vonJMC5053
- INFILTRACIÓN CHINECASHochgeladen vonRandi Bolaños
- Flow Chart of Pouch, Cylinder and Coin Cell by MTI's Battery Equipment (1)Hochgeladen vonhca
- A2 PracticalsHochgeladen vonMushfique Ahmed
- Sterilization Validation QsiteHochgeladen vonskype2121
- Journal of Ornamental and Horticultural Plants.pdfHochgeladen vonManu Manda
- Aqua Cell 4Hochgeladen vongokmenozkan
- Gulf Per Lite Loose Fill Insulation TdsHochgeladen vonilieoniciuc
- 1. Outline ReportHochgeladen vonsygwapooo
- Ecologyecosystems 150313032615 Conversion Gate01Hochgeladen vonPrakriti Goel
- 22 Condensors & EvaporatorsHochgeladen vonPRASAD326
- Comparison of Fuel Economies of High Efficiency Diesel and HydrogenHochgeladen vonndlongnd
- Marine Grease 3121 TDSHochgeladen vonsatyendra
- Max Allowable Flange LoadHochgeladen vonprasoon
- GB std steel.pdfHochgeladen vonBapu612345
- 3D printing.pdfHochgeladen vonDiego Martínez
- 3. Tyfo WebHochgeladen vonDediMulyadi
- Sun Path DiagramHochgeladen vonKiara Alejandrino
- Sci 207 Week 3 Assignment Outline if the Final Lab Report – NewHochgeladen vonjohn
- MtDNA ExtractionHochgeladen vonJanani S
- Properties of insulating materials.pdfHochgeladen vonLincoln Cesar
- Suggested Answer_tutorial 3 Chm510 (Mac2017)Hochgeladen vonMark Sullivan
- Use of Rapid Prototyping in Rapid ToolingHochgeladen vonSudhanwa Kulkarni
- Quantum Cascade LaserHochgeladen vonNit Gossy
- 0 - Introduction to DIANA.pdfHochgeladen vonMaría Elena Novelo Arjona
- Carew Et Al.mite PopGen.entexpAppl.2004Hochgeladen vonjeanlouiskouakou
- Nitrogen Syngas Vienna Austria 2006 Methanol Casale and Foster WheelerHochgeladen vonGeorge Van Bommel
- Fire Characteristics of Styropor FoamsHochgeladen vonRosman Al-Qadri
- Generating McCabe-Thiele Diagrams Using HYSYSHochgeladen vonAhmed Mohamed Khalil
- Main ProjectHochgeladen vonseelam manoj sai kumar

## Viel mehr als nur Dokumente.

Entdecken, was Scribd alles zu bieten hat, inklusive Bücher und Hörbücher von großen Verlagen.

Jederzeit kündbar.