Comparison of Methods For The Measurement of Fibre - Matrix Adhesion in Composites

Comparison of methods for the measurement of fibre/matrix adhesion in composites
P.J. HERRERA-FRANCOand L. T. DRZAL (Michigan State University, USA) Received 14 October 1991; accepted 16 October 1991
This paper reviews the most common state-of-the-art techniques--microbond, single-fibre fragmentation and microdebond/micro-indentation--for measuring fibre/matrix adhesion in composite materials. Following a discussion of the background for each technique and theoretical analysis, the experimental apparatus and procedure are described. Results of finite element and photoelastic analyses are presented to identify the state of stress that is induced in each specimen model of the three techniques. An objective comparison between the three single-fibre techniques to measure interfacial shear strength, and between three composite laminate techniques ([+45] s tensile, Iosipescu and short-beam shear), is made using AS-4 and AU-4 carbon fibres in an Epon 828 epoxy resin system as the experimental material. Finally, the advantages and limitations of each technique are discussed.
Key words: composite materials; fibre~matrix interface; adhesion; singlefibre techniques; microdrop; fibre fragmentation; micro-indentation; interfacial shear strength
It is well known that the level of adhesion between fibres and matrix affects the ultimate mechanical properties of a composite, not only in the off-axis direction but also parallel to the fibres. Several models which relate structure and property to composite fibre/matrix interfacial behaviour have been proposed. These are based either on mechanical principles with some assumptions made about the level of fibre/matrix adhesion in the composite, or have taken a surface chemistry approach in which the interface was assumed to be the only factor of importance in controlling the final properties of the composite. Neither effort has had much success. A growing body of experimental evidence points to the existence of a region different in structure and composition near the fibre/matrix interface. These results have led to an understanding of the inter-relationships between fibre, interface and matrix, giving birth to the concept of an interphase; i.e., a three-dimensional region existing between bulk fibre and bulk matrix ~. This interphase includes the two-dimensional region of contact between fibre and matrix (the interface) but also incorporates the region of some finite thickness extending on both sides of the interface in both the fibre and matrix. The complexity of this interphase is illustrated schematically in Fig. 1. It has also been
s h o w n 1-9 that coupling the mechanics and the surface chemistry approaches, the structure and composition of the fibre/matrix interphase as well as the state of stress, results in a better understanding of the composite static and dynamic final properties.
There have been many attempts to develop a technique which could measure fibre/matrix adhesion levels in composites with high fibre volume fraction. This would allow evaluation of the processing or environmental exposure encountered either during manufacturing or in service. Moisture, solvent sorption, fatigue and thermal exposure could then be properly evaluated for their effect on the fibre/matrix interface and composite properties. The experimental methods used to characterize the fibre/matrix interphase during recent years have relied on the use of single fibre-matrix adhesion and failure modes. The oldest technique is the fibre pull-out method m which was developed in the early stages of composites research when the fibres were much larger and easier to handle. Variations in the details that have evolved mostly involved the matrix portion. The fibre may be pulled out of the matrix which can be a block of resin, a disc, or a droplet ~~. Current trends include the use of very small droplets in order to reduce the diffi-
0010-4361/92/010002-26 (~) 1992 Butterworth-Heinemann ktd 2 COMPOSITES. VOLUME 23 . NUMBER 1 . JANUARY 1992
Bulk matrix
Polymer of different properties
!:~L]:!!:~i
Thermal, chemical, mechanical, environments
Fibre topography
I
Fibre morphology Bulk fibre
Fibre/matrix interphase Fig. 1 Characteristics of the fibre/matrix interphase in a composite material
culties in preparing thin discs of resin and to reduce the variability in exit geometry. Another method uses a single fibre totally encapsulated in a matrix coupon, which in turn is loaded in tension. The interracial shear stress transfer mechanism is relied upon to transfer tensile forces to the encapsulated fibre through the interphase ~2"13.The fibre tensile strength (of) is exceeded and the fibre fractures inside the matrix. This process is repeated producing shorter and shorter fragments until the remaining fragment lengths are no longer sufficient in size to produce additional fracture through this stress transfer mechanism, and a shear-lag analysis is completed on the fragments in order to calculate the interfacial shear strength.
A n in situ measurement technique has been proposed for measuring the fibre interfacial shear strength 14. It involves the preparation of a polished cut surface of a composite in which the fibres are oriented perpendicular to the surface. A small probe is placed on an individual fibre and the force and displacements are monitored in small steps to the point at which the fibre detaches from the matrix.
followed with increasing load. The fracture energy of the interface was calculated as a function of the strain in the resin, the tensile modulus of the fibre, the frictional shear stress (~), the length of the debond (x) and the fibre diameter (a). Narkis et al. 16proposed the use of a single-fibre specimen in which the fibre is embedded along the centre line in the neutral plane of a uniform crosssection beam. The beam is placed in non-uniform bending according to an elliptical bending geometry. This causes the shear stress to build up from the end of the fibre according to the gradient of curvature of the specimen. Careful observation of the fibre allows location of the point at which the fibre fails because of a maximum shear stress criterion. The stress along the fibre is calculated as a function of the matrix tensile modulus (Era), the beam width (b), the first moment of transformed cross-sectional area (Q), and A and B which are constants from the equation of the ellipse. Ko et al. 17examined a carbon fibre/epoxy system in which the interfacial properties had been varied by the use of dynamic mechanical analysis (DMA). They report a change in the tan 6 peak attributable to changes in the fibre/matrix interface. Chua 18also measured a shift in the loss factor for glass/polyester systems that corresponded to changes in the condition of the fibre/matrix interface. Perret et al. ,9 measured both the loss factor and the change in the shear modulus with increasing displacement and detected a change in composite
Less direct methods for measuring the interaction between fibre and matrix have also been reported. Outwater and Murphfl 5 used a single fibre aligned axially in a rectangular prism of matrix. A small hole was drilled in the centre of the specimen through the fibre. The prism was placed under a compressive load and the propagation of an interfacial crack was
COMPOSITES. JANUARY 1992
properties with a change in the fibre/matrix interface. Yuhas e t al. 2 have used ultrasonic wave attenuation to establish correlations with short-beam shear data. This method was useful for poorly bonded systems but was not sensitive to well-bonded interfaces. Recently, localized heating has been coupled with acoustic emission events to detect interfacial debonding 21. Correlations were made between acoustic parameters and interracial strength. The ability to measure fibre/matrix interfacial properties, such as modulus of the interphase material, will be necessary for predictive models that couple interfacial properties to composite properties. Further development of interphase-sensitive techniques will be forthcoming. This paper reviews the most common state-of-the-art techniques, focusing on the methods of assessing the fibre/matrix mechanical interactions, and addresses the theoretical analyses upon which these methods are based. Also, finite element and photoelastic analyses were performed in order to identify differences in the state of stress that is induced in each specimen model of three different techniques. In order to provide an objective comparison between the three different techniques to measure the interfacial shear strength, an AS-4 and AU-4 carbon fibre/Epon 828/meta-phenylene diamine (mPDA) epoxy resin system was selected as the experimental material. The AS-4 fibre exhibits significantly higher adhesion to this matrix than the AU-4 fibre.
Load, F
3 2r
2R
Load,F Microvise
\ Microdroplet Fibre diameter (d)
b
Fig. 2 Schematic representation of the fibre pull-out and microbond techniques showing geometrical parameters: (a) after Piggott; (b) after Miller
MICROBOND TECHNIQUE Background

The pull-out experiment, which is believed to possess some of the characteristics of fibre pull-out in composites, consists of a fibre or filament embedded in a matrix block or thin disc normal to the surface of the polymer (Fig. 2). A steadily increasing force is applied to the free end of the fibre in order to pull it out of the matrix m. Load and displacement are monitored as the fibre is pulled axially until either pull-out occurs or the fibre fractures. The strength of the fibre/matrix interface can be calculated to a first approximation by balancing the tensile stress (of) on the fibre and the shear stress (T) acting on the fibre/matrix interface, obtaining a simple relationship of the form = ( o f / 2 ) ( d / L ) , where it is assumed that the shear stress is uniformly distributed along the embedded length and d is the fibre diameter l0 . Several theoretical models have been proposed to determine the shear stresses developed during pull-out. Greszczuk 22 considered the case of an elastic matrix in which the shear stress distribution was no longer uniform and the load transferred between the fibre and matrix did not change uniformly along the fibre. He showed that distribution of stresses and forces depend on the properties of the elastic matrix. Lawrence :3 further developed Greszczuk's theory by including the effect of friction. Takaku and Arridge 24 considered the effect of the embedded length on the debonding stress and the pull-out stress, and also the effect of Poisson's
contraction on the variation of pull-out stresses. Gray 25 reviewed and applied the previously mentioned theories 23"24to calculate the maximum shear stress when the fibre is pulled out from the elastic matrix. He concluded that the mixture of adhesional bonding and friction resistance that occurs in a pull-out test specimen depends on the length of the embedded fibre. Adhesional bonding increases with embedded fibre length, whereas the frictional resistance to pull-out due to friction decreases. Laws 26 calculated the load/ displacement curve of a pull-out test based on Lawrence's theory 23, the crack spacing and strength of an aligned short fibre composite, and the effect of the interfacial and frictional bonds on pull-out. More recently, Banbaji 27"z8presented a theoretical model that considered the effect of normal transverse stresses on the pull-out force. He first analysed the case in which the normal stress is Constant, and then the case in which the stresses depend on the way the tensile force changes during an actual test. He applied the results to a polypropylene-cement system. A variation of the pull-out technique was reported recently by Hampe 29. One end of the fibre is embedded in a small amount of polymer in the form of a hemisphere formed on the surface of a metal plate. The sample is mounted on an electronic balance capable of measurements with an accuracy of +0.1 mN. The free end of the fibre is clamped in a motor-driven support capable of speeds from 5 ~m min- 1 to 5 mm min- 1 . The displacement is measured with an accuracy of
+0.3 ~tm. Data acquisition and control of the apparatus is done using a computer.
Fd=2~rLziy
(6)
Theoretical considerations 1. Pull-out technique

In this paper theoretical considerations presented by Chua and Piggott 3-33 will be described. They re-examined the pull-out process and showed that it is governed by at least five different variables: interracial pressure (p), friction coefficient (~t) along the debonded length, work of fracture of the interface (Gi), the embedded fibre length (L) and the free fibre length
3) It has been observed from experimental results that the mode of failure is sudden and catastrophic. This type of failure is more cummon to brittle fracture than a maximum stress criterion. Since a stress concentration exists at the point where the fibre emerges from the matrix, this is the most likely point where failure probably starts and propagates rapidly along the interface. Failure may occur if the interface fractures with work of fracture (Gi) per unit area of interface. The source of the required fracture surface energy is the strain energy stored within the specimen components 31-42. Chua and Piggott 3'34 considered only the extensional strain energy stored in the embedded fibre length (UL). The shear strain energy in the matrix immediately surrounding the fibre (Um) is given in the following equation, where n is given by Equation (2) and s=L/r:
(10.
Assuming that both fibre and matrix behave elastically and that at the interface stress transfer occurs from matrix to fibre without yielding or slip (perfect bonding), they developed a relationship for the tensile stress within the fibre (of) at any point along the embedded length, based on previous work by Greszczuk and Lawrence22"23:
nr~o~ecoth(ns)
ULq'- U m (7)
sinh[n( L-x)/r]
(If = (3 fe
(1)
2nEf
Equating the total strain energy to 2nLG c, where G c is the unit fracture energy at the interface, the debonding load Pd is found to be:
sinh(ns)
where s=L/r, L is the length of the embedded fibre, r is the radius of the fibre and ofe is the fibre stress at the polymer surface, that is, the average tensile stress on the fibre. The geometric terms are also shown in Fig. 2 and
n2 = Em
Pa =2nrX/EfGcr(ns)tanh(ns)
(8)
(2)
Ef(1 + vm)ln( R/r ) where Ef and E m are Young's moduli for fibre and matrix respectively and Vm is the matrix Poisson's ratio. The shear stress at the interface is calculated from the equilibrium of forces exerted on a differential fibre element, of length dx, to give the well-known equation:
~i ~--~-
Penn and Lee 35 considered the existence of an initial microcrack of length a at the fibre/matrix interface. They also considered the effect of the strain energy contributed by the free fibre length to the crack propagation process, and using an energy balance they derived an expression for the debonding force Pd of the microdrop: 2rtrX/~rGcEf (9)
(rtdf ~ "~
Pal= [l+csch2(ns)]
After debondin~, friction at the interface has to be overcome 24 "27 2~.32 i n o r d e r f o r pull-out toproceed. " Friction at the interface is due to the normal compressive stresses that are caused by the pressure Po acting on the fibre from the matrix. Such stresses xi= ~tPo (~t is the coefficient of friction) arise from resin shrinkage during curing of the specimen and from the difference of coefficients of thermal expansion of matrix and reinforcing fibre 27'32"43. Interracial shear stress, ~i= ~P, increases with increasing pull-out distance 36, but shrinkage of the matrix and Poisson's effects on the fibre due to the applied tensile force will produce a reduction of fibre cross-sectional area relative to the matrix, resulting in a reduction of both interfacial normal and shear stresses. In a typical force/displacement plot (Fig. 3) describing the pull-out process, the first peak is attributed to debonding and frictional resistance to slipping, whereas subsequent lower peaks are attributed to friction coupled with a stick-slip mechanism, giving rise to a serrated appearance of the curve 16"24'27'2s. Because of relaxation at both the free and embedded lengths of the fibre, the slope of a curve (x~) against pull-out distance gives only an approximate value of the interracial shear stress (~tpo) where Po is the pressure exerted by the
(3)
which, using Equation (1), results in:
cosh[n(L-x)/r]
17, = n ( i f e i
2sinh(ns)
(4)
It should be pointed out that no bonding is considered across the end of the fibre. During a pull-out experiment there are three possible failure modes of the fibre/matrix interface 34. 1) Failure may occur when the principal shear stress reaches the interface strength (~iu), which has a maximum absolute value at x = 0 , that is, at the polymer surface. The debonding force is then Fj=nrZ(ife. From Equation (4): Fd =
2nrZxiutanh (ns )
n
(5)
2) Yielding at the interface might also occur if its yield strength Tiy is reached, in which case a constant shear stress distribution can be assumed along the embedded fibre length, as long as work-hardening effects are negligible, and thus:
COMPOSITES
. JANUARY
1992
1.5
Table 1. forces 2s
Effect of pressure on m a x i m u m pull-out
Pressure (MPa)
1.0
Embedded Interfacial length Pull-out force shear strength (mm) (N) (MPa) 4.20 4.30 41.2 62.0 1.20 1.20
~9 t~ o9 0.5
10mm
I mm
17.6 35.2
0.2
0.4 (ram)
0.6
Pulled-out distance
Fig. 3 ment
Typical force/displacement plot from a pull-out experi-
matrix shrinkage at the moment the fibre emerges from the polymer. The experimental value for shear stress (~ep) obtained from the slope of the pull-out curve is related to the true value of Ti= ~Po by Equation (10):
"~exp
(10)
1;i--[ 1-{'- 2*~xp(I+2L) ]Efr

If, in addition to curing stresses, an external pressure Pe is applied, the shear stress increases as: i= ~t(p~+p0) (11)
The value of ~tcan be determined from experimental results by plotting T~against p~ and evaluating the slope of the curve. The intersection of the curve with the vertical axis should give ~tpo. If no external pressure is applied, ~ can be estimated from the curve of the pull-out force F A as a function of the pull-out distance: :gd2Ef(1 +Vm) FA= vfEm(1, e nenx/L ) (12)
distributed along the embedded length of the fibre 1O. The average shear stress is calculated by dividing the maximum measured force of debonding by the embedded fibre length area. Using photoelastic and finite element analysis, a study of the state of stress along the embedded length of a fibre in a microdrop was conducted45. Since the stresses are not constant through the thickness of the specimen, the stress values obtained from the photoelastic analysis are actually an average effect. Thus a correction to obtain the shear stresses (or principal stress difference) was necessary and the analytical form has been reported elsewhere45.46. First, the ideal case of a perfectly spherical droplet was analysed. In a recent study47, it was shown that a change in the shape from a spherical droplet to an elliptical droplet does not affect significantly the state of stress along the embedded length of the fibre. A second geometry considered the presence of a meniscus formed by the matrix at its point of contact with the fibre, and its effect on the interfacial stress distribution and failure mode. During an actual experiment the matrix is held fixed against an adjustable stop positioned so that the fibre can pass. In the finite element simulation, the effect of changing the relative position of the point of contact with the microdrop on interracial stress distribution is considered. At the point of contact a stress concentration exists at the two diametrically located contact points, but the state of stress around the fibre is not axisymmetric. Only the plane which contains the two points of contact was considered. Thus, for both models, a state of plane stress is assumed. Fig. 4 shows isochromatic fringe patterns obtained from photoelasticity experiments and finite element modelling for a spherical microdrop. Figs 5 and 6 show the interracial shear stress distributions for different loading conditions obtained from the photoelastic and finite element analyses respectively. In this case the stresses are normalized by the applied fibre tensile stress, position and the diameter of the fibre. The shear stress distribution obtained from photoelasticity corresponds to a continuous support uniformly distributed along the upper portion of the photoelastic model. In the finite element analysis, several positions of the point of contact were considered together with the continuous support. All these cases were obtained for the same load applied to the free end of the embedded fibre. It can be noticed that when straight supports are used to support the droplet, the peak shear stress is higher as the point of contact gets closer to the fibre while the uniformly distributed support gives a peak value intermediate to the results from the support positions considered here. The position of occurrence
where vf is the fibre Poisson's ratio, ~l=2~tvEmL/ Efr(l+vm), and x is the pulled distance. The effect of the external pressure applied to the specimen is more noticeable on the pull-out force at short embedded lengths27"44. It can be seen in Table I that, for similar values of embedded fibre length, doubling the external pressure increases the pull-out forces. However, this increment is offset by the decreasing thickness of the fibre due to both the pressure and lateral Poisson's contractions that are produced upon application of the initial tension, which results in the decrease in the coefficient of friction.
2.
Microbond technique
In the case of the microbond or microdrop technique, it is assumed that the interracial shear stress is uniformly
0.20
0.15
\
O
\
O
~q
0.10
~J tu~
\ \
0.05
\
o
r-
\
0
\ \
\
0.00 0
I I
e~o
1 I
'o~
8
Fig. 4 Isochromatic fringe pattern obtained from: (a) photoelastic analysis; (b) finite element analysis for a microbond model. Insert shows studied area
x/d
Fig. 5 Interfacial shear stress distribution along the fibre/matrix interface in the microbond technique obtained from a photoelastic analysis corresponding to blade support and continuous support
of the maximum shear value is also dependent on the position and type of support used. As the gap between the supports becomes wider, the peak stress position shifts towards the centre of the microdrop. It is also noticed that there exists a small oscillation of the shear stresses close to where the fibre emerges from the microdrop. This oscillation could be attributed to the stress singularity existing at the fibre/matrix interface due to the sudden change in material properties 48. However, the peak stress is not affected by this singularity and is only caused by the type of loading or support used during the experiment. Fig. 7 shows the corresponding radial stresses for the same loading conditions, obtained from the finite element analysis. In this case or is tensile at the point where the fibre emerges from the droplet, decreases rapidly towards zero and then becomes compressive, going through a peak value which decreases to low values at points located closer to the centre of the drop. Again it is noticed that the loading conditions affect the radial stress distribution both in magnitude and point of occurrence. These tensile radial stresses could play an important role in the failure initiation at the interface. Thermal stresses in the microdrop are not uniformly distributed along the interface due to changes in the geometry of the microdrop, thus making the radial stresses arising from the applied loads very important.
It should be noticed that differences as small as 4 ~tm in the relative position of the point of contact of the supports with the microdrop (expressed as an angle formed by the two points of contact and the centre of the drop) result in high differences in both the peak shear stresses and their location. This factor may be responsible for the high scatter usually observed with experimental microdrop results. In a typical experiment, the spacing of the supports is always changed from specimen to specimen and adjusted to a gap just wide enough to allow the fibre to move between them. From one drop to another there will be differences in spacing resulting in differences in the relative point of contact and consequently to changes in the state of stress at the fibre/matrix interface. Since in this experiment only the magnitude of the debonding force is measured, depending on the point of blade contact, the interface strength might be reached at different values of applied force for drops of the same size and consequently different interfacial strength values will be calculated. When drops are formed on the fibre, they conform to its cylindrical shape but, depending on the contact angle and volume of the drop, a meniscus will always exist at the contact point of fibre and matrix49. The presence of the meniscus will make the measurement of
0.225
'
~
--v-v-
ip T
i
'
'
0.175
<
0 2
Tt
p
I
0.30
0.25
0.20
~01< 02
,~ 0.150 /
OlOO ,o,,o /
0~02
y
0.15
/./
\
X
"~L 0.10
I.
-O--O-
o-,,,,.
"X~Ox~x.
PT Y/
0"075~//~ /
0.050 0.025
0.05
\
X7
o\,\
o v
o.oo, / \
-005
\
J t
o/, /
/'o...o/
0.000
O.
0.5
1.0
I .5
2.0
x/d
2.5
3.0
3.5
4.0
-o.lo
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 50
x/d
Fig. 6 Interracial shear stress distribution along the fibre/matrix interface in the microbond technique obtained from a finite element analysis including blade and continuous supports
Fig. 7 Interfacial radial stress distribution along the fibre/ matrix interface in the microbond technique obtained from a finite element analysis including blade and continuous supports
the embedded length of the fibre very inaccurate. As can be seen in Fig. 8, the meniscus does have an effect on the shear stress distribution along the fibre/matrix interface. The oscillation of the shear stress and the peak stress value are higher, and dependent on the different loading conditions. Radial shear stress (not shown here) follows the same trends as in the case of the spherical drop with the exception of higher oscillations. Fig. 9 shows the locus of maximum tensile stresses and hence the locus of possible tensile failure in the direction of the applied force as a heavy line. Upon debonding it is very common to observe that, at the initial position of the droplet on the fibre, a small cone of matrix material is left behind. This could be attributed to the tensile stresses in the meniscus area. As a result, debonding due to premature failure of the matrix could trigger interfacial failure, thus resulting in a lower value of debonding force.
thermosetting resin matrix. The procedure used in this test and outlined here is similar to those described by other investigators 5(r-57. 1) The ends of 100 mm long fibres are taped to parallel sides of a specially built frame using double-sided tape. 2) After mixing and degassing the resin, microdrops are applied to the fibres using a syringe and needle. A small drop of the thermoset resin is made to flow to the needle tip and is allowed to come in contact with a fibre After retraction of the needle tip, some of the resin remains with surface tension forces forming a microdrop around the fibre The microdrop size ranges from 80 to 200 pm in diameter 3) The microdrops are allowed to cure at the processing conditions selected for the matrix materials. 4) The fibre is affixed to an aluminium plate and kept in a desiccator to await testing. The aluminium tab has dimensions 6 x 15 x 1 mm and a hole 0.8 mm in diameter is drilled in one end for connecting to the load cell. 5) To test the microdrops, the small aluminium plate is attached to a load-sensing device. The droplet is gripped with micrometer blades which are brought together until they nearly touch the fibre. Before loading, the fibre diameter and embedded lengths
Experimental apparatus and procedure

Fig. 10 shows a photograph of a typical microdroplet pull-off test apparatus. The following experimental procedure can be used to measure the interracial shear strength by means of the microbond technique for a
COMPOSITES.
JANUARY
1992
0.35
I P
-'0-'-0-O. 30
C~
0.25
0.20
- ~
$22 Value
~':i~iii:il;;i
-IR--I
1-2.s2E-o3:::!!!!!!!:!!!!!:!!: i:::N:I:I::::!;i i i i i i i !i i
.63E-03 !i!!!!!: ~!~!i!!
0.15
-~N--7.42E-O4 !iiiiiiiiii!
~-2.97E-04
~+1.47E-04
~iiiiii ~i'iii~ii:!:ili ..........
Locus
0.10
i!iiii~iiiiiiii ilili!iiii!iiiii!i
;,........ :Z:T
fai,ure
iiii[iiiiiiii!iiiiiii!i!i!i!iii!!i!!i!!
i:i~i~i~i!i{i{i........ ;!i! ~iiii!ili!iiii!ilill ..................
0.05
iiiiiiiiiiiiiiii!i!i i!iiii!iiiiiiiiiii
i!i!i!i!!!!!!!ilili~iii~iiiii~iii!i!i!i!
i!i!ii!ii!i~ii{iii!i:iiiNi!f~fii~i!ii!!i!i!i?!!i!ii i
i~i~ili~ii~i~!~!!i! ..................
~o~
0.00 I I I I .5 I I f
0.0
0.5
I .0
2.0
x/d
2.5
3.0
3.5
4.0
Fig. 8 Effect of a meniscus on the interfacial shear stress distribution along the fibre/matrix interface for the microbond technique
Fig. 9 Locus of tensile failure across the meniscus formed in a microdrop (insert shows area of study)
are measured using a Cue Micro 300 Digital Video Caliper (Olympus Corporation). The micrometer blade, mounted on a translation stage, is moved away from the load cell parallel to the longitudinal axis of the fibre, at a maximum speed of 22 Ixm min-L The position and speed of the translation stage are controlled remotely using a motorized actuator. The translation of the stage causes the microdrop to be sheared off the fibre surface. The force required to debond the microdrop and debond length are recorded using a computer, and an average interracial shear strength is calculated. A method to form microdrops of a thermoplastic matrix material usually consists of the following steps. 1) Carefully remove fibres from the tow and place them on a rectangular open frame with double-sided tape. Store the fibres in a desiccator until ready to form the microdrops. 2) Cut thermoplastic sheet into small squares approximately 3 mm to 5 mm per side and place them in a beaker containing an appropriate HPLC grade solvent at a concentration of approximately 5% by weight. Seal the container and agitate the mixture overnight. When mixing is completed, keep the solution in a tightly sealed container in a refrigerator to minimize evaporation.
3) Apply the thermoplastic/solvent solution by drawing along the fibre a small drop of liquid formed at the tip of the disposable pipette. In order for a drop to remain on the fibre it is necessary to wait until the small drop at the tip of the pipette turns cloudy, that is, when some of the solvent has partially evaporated. The size of the final drop is controlled by the speed at which the liquid is deposited on the surface of the fibre. If larger drops are required, four or five passes might be necessary. Leave the drops overnight in a forced ventilation hood to allow the solvent to evaporate. 4) Place the drops in a forced convection oven. It is important to keep the frame and the fibres in a vertical position to maintain the concentricity of the drops on the fibre. Quickly ramp the temperature of the oven to a point greater than the Tm of the matrix and maintain this temperature for 40 min to 1 h. This will allow any residual solvent to evaporate. Bring the oven back to room temperature as quickly as possible. 5) Cut the fibre into 50 mm long segments and attach one to an aluminium tab with cyanoacrylate adhesive. The aluminium tab has the dimensions 6 15 x 1 mm and a hole 0.8 mm in diameter is drilled in one end for connecting to the load cell. It is important to ensure that the edge of the tab in
already affixed to a holding frame, and the thermoplastic is melted on the fibres. Upon melting, nearly uniform sized droplets are obtained. Their diameters are controlled by the film thickness.
SINGLE-FIBRE FRAGMENTATION TECHNIQUE Background

Kelly and Tyson 12first used the single embedded fibre tensile specimen, in this method a fibre is embedded in a matrix material such that the strain to failure of the matrix is at least three times higher than that of the fibre. They observed a multiple fibre fracturing phenomenon upon application of a tensile force to a system consisting of a low concentration of brittle tungsten fibres axially aligned and embedded in a copper matrix. As the applied strain increases, the embedded fibre breaks repeatedly at points where the fibre strength (or) has been reached. Continued application of strain to the specimen results in repetition of this fragmentation process until all remaining fibre lengths become so short that the shear stress transfer along their lengths can no longer build up enough tensile stresses to cause any further failures (Fig. 12). This maximum final fragmentation length of the fibre is referred to as the critical length, 'lc'. The shear stress at the interface is assumed to be constant along the short fibre critical length (also assumed to have a constant diameter). An average shear strength (x) can be determined from a simple balance of force which results in:
Fig. 10 Apparatus of the microbond technique showing detail of blade shape: l - - l o a d cell; 2--blade micrometer; 3 - - X - Y translation stage; 4--actuator; 5 ~ b a s e
contact with the fibre does not have any sharp corners in order to avoid premature fibre failure. Keep the samples in a desiccator until they are ready to be tested. Testing is performed using the same procedure outlined in step (5) for thermosetting resin microdrops. A different method to form microdrops for various fibre/thermoplastic systems, illustrated in Fig. 11, has been reported by Gaur et al. 51. This procedure was used to measure the interfacial shear strength of carbon and aramid fibres embedded in four thermoplastic resins: polyethetherketone (PEEK), polyphenylene sulphide (PPS), polycarbonate (PC) and polybutylene terephthalate (PBT). The procedure is as follows: 1) A longitudinal cut is made in a small, thin piece of film (about 2 to 30 mm) along nearly its entire length, to form two strips joined at one end for a distance of 50 to 100 ~tm. The specimen has the appearance of a pair of trousers. 2) The specimen is suspended on the horizontal fibres
(J
m immm ~
L,___J
o3 > o2 > o1 /c
T. I
Heat
+~ Thin strip of polymer film

(cut down centre and then spread apart) P o l y m e r film 2. !
Fig. 11 fibre sl
Procedure to form thermoplastic resin droplets on the
Fig. 12 Schematic representation of the single-fibre fragmentation process
10
COMPOSITES . JANUARY 1992
= -2
(13)
where d is the fibre diameter for a circular fibre cross-
section. Since the fibre/matrix interface is placed under shear, the calculated value of'~ is often used as an estimator of the composite interfacial shear strength 13. The distribution of stress around discontinuous fibres in composites has been studied by a number of researchers. Theoretical analyses have been performed by Cox58 and Rosen 59. In these models only fibre axial stress distribution and the fibre/matrix interfacial shear stress distribution are determined. Amirbayat and Hearle 6 studied the effect of different levels of adhesion on the stress distribution; that is, no bond and no adhesion, perfect bond and the intermediate case of limited friction. They also considered the inhibition of slippage by frictional forces resulting from interfacial pressure due to Poisson's lateral contractions of the matrix but did not consider the shrinkage of the matrix during curing. Theocaris61 proposed a model that incorporates an interphase which he named a mesophase, which constitutes a boundary layer between the main phases of the composite. A continuous and smooth transition of the properties from one phase to the other is assumed. Because the mechanical properties of this region also contribute to the composite properties, the determination of the local mechanical modulus is important. Dynamic mechanical analysis was used to identify the mesophase properties, primarily the glass transition temperature (Tg), through changes in the loss modulus peak. The effect of adhesion as a function of the surface treatment on the fibre, and the ratio of elastic moduli of fibre and matrix and temperature on the critical aspect ratio was analysed experimentally by Asloum e t al. 62 . Folkes and Wong 63 , in their study of adhesion between fibre and matrix of thermoplastic composites, noticed that the formation of transcrystalline morphology around glass fibres in polypropylene has an effect on the critical fibre length, probably through the change in local interphase modulus. Lhotellier and Brinson 64 developed a mathematical model that includes the mechanical properties of the interphase, the stress concentration near fibre breaks, and the elasto-plastic behaviour of both the matrix and the interphase. Rao and Drza165 and Drza166 studied the dependence of the interfacial shear strength on the bulk material matrix properties using model compounds based on epoxy/amine chemistry. AS-4 carbon fibres were used as the subject for these measurements with both a difunctional epoxy (diglycidyl ether of bisphenol A, DGEBA)system as well as a tetrafunctional epoxy (MY720) system. In order to produce matrices with a range of matrix properties from brittle elastic to ductile plastic, primary amine curing agents with varying backbone compositions were carefully selected. The fibre/matrix interfacial chemistry was kept constant throughout this study by always using the same amount of curing agent. They found that for the difunctional as
well as the tetrafunctional epoxy system, the interfacial shear strength (measured using the single-fibre fragmentation technique) decreases non-linearly with decreasing modulus of the matrix. Linear elastic analysis yields a nearly linear relationship, for both systems, between interfacial shear stress and the product of strain to final break and the square root of the matrix shear modulus. A linear relationship is also found between the difference in test temperature and glass transition temperature of the cured matrix and the interfacial shear strength. Additionally the failure mode is seen to remain interfacial as the ductility of the matrix changes. Termonia 67 used a finite-difference approach to show that the critical length for efficient stress transfer to the fibre is a function of the ratio between elastic moduli of the fibre and matrix. In his model he also considered the dependence of the critical length on the adhesion by including an adhesion factor. A decrease in the adhesion is seen to increase the critical length, particularly when the adhesion factor becomes less than 30%. In a recent study, Galiotis e t al. 6s used Raman spectroscopy to determine strain profiles along the fibre fragment length on surface-treated and non-treated carbon fibre. They reported that, for the treated fibre, debonding at the crack tip initiates at the fibre fracture strain. The maximum interracial shear stress per increment of load is obtained at a certain distance from the crack tip which is equal in size to the combined debonding and matrix yielding zones. The maximum interfacial shear strength profiles for the non-treated fibre indicated that the load transfer between fibre and matrix is obtained through friction only. Verpoest et al. 69 recently presented a micromechanical analysis concluding that, in a few special cases, the critical fibre aspect ratio can reach values which are lower than those predicted by Kelly's shear-lag analysis. They also proposed that this method could be used to estimate the different components of the interface shear strength, i.e., the bond strength, the friction strength and the matrix yield strength.
Theoretical considerations
Whitney and Drzal presented a theoretical model to predict the stresses in a system consisting of a broken fibre surrounded by an unbounded matrix 7. The model (Fig. 13) is based on the superposition of the solutions to two axisymmetric problems, an exact far-field solution and an approximate transient solution. The approximate solution is based on the knowledge of the basic stress distribution near the end of the broken fibre, represented by a decaying expo-
Fig. 13 Micromechanical model ofthe single-fibre fragmentation test TM
C O M P O S I T E S . J A N U A R Y 1992
11
nential function multiplied by a polynomial. Equilibrium equations and the boundary conditions of classical elasticity theory are exactly satisfied throughout the fibre and matrix, while compatibility of displacements is only partially satisfied. The far-field solution away from the broken fibre end satisfies all the equations of elasticity. The model also includes the effects of expansional hygrothermal strains and considers orthotropic fibres of the transversely isotropic class. A relationship is obtained for the axial normal stress Ox in the fibre (see Fig. 13): Ox = [1- (4.75+1)e-475~]C1% (14)
(73= and
Elf +
4KfGmvl2f(Vl2f--Vm) (Kf+Gm)
(22)
gf=
2 ( 2
Em
E2f
2v22fE2f~ (23)
2G2f
Elf ,/
where the thermal strains are indicated by overbars, and E2e, G2f and Kf are the radial elastic modulus, the shear modulus in the plane of the cross-section and the plane-strain bulk modulus of the fibre, respectively. A numerical example is presented for a single-fibre composite of AS-4 carbon fibre/Epon 828 epoxy matrix cured with a stoichiometric amount of mPDA (Aldrich Chemical). The material properties are shown in Table 2. The specimens were cured at 75C and postcured at 125C. The difference between room temperature, 21C, and the post-cure temperature yields A T = 104C, which is the worst case for residual thermal stresses. Because it is most likely that some residual stresses will be relieved during cool-down from the post-cure temperature, in this example the value A T=75C was chosen. Figs 14, 15 and 16 show plots of Oxf/Oo,~rx/o,, and oJoo, where the fibre far-field stress was used to normalize out the numerical results. It can be seen that the axial fibre stress and the interracial shear stress are relatively insensitive to thermal strains, but the radial strain is quite sensitive to thermal strains. As noticed by other investigators, the location of the maximum shear stress is at some distance from the broken end of the fibre 71-73.
where =x/lc, Co is the applied far-field strain and C1 is a constant dependent on material properties, thermal strains and the applied far-field strain. It can be noticed that o x is independent of the fibre radius. The critical length Ic is defined such that the axial stress recovers 95% of its far-field value, that is: ox(lc) = 0.95C1% (15)
where (O<.r<.R). The interfacial shear stress is given by the expression:
1;xr= --4.75~61coe-4'75~
where
(16)
(O<.r<-R) and:
~'E f-"4-"~lfGm am 1 2
(17)
E~f denotes the axial elastic modulus of the fibre, whereas vlaf is the longitudinal Poisson's ratio of the fibre, determined by measuring the radial contraction under an axial tensile load in the fibre axis direction, and G m denotes the matrix shear modulus. It should be noted that the negative sign in the expression for the shear stress is introduced to be consistent with the definition of an interfacial shear stress in classical elasticity theory. The radial stress at the interface is given by: or =
Photoelastic and finite element stress analysis

There exist in the literature several reports on the experimental determination of the shear stress distribution around a discontinuous fibre in a model composite material 46'74'75. The results presented in these studies relied on two-dimensional models simulating discontinuous fibres embedded in a matrix. A typical isochromatic fringe pattern obtained from the photoelastic analysis and one generated from the finite element results are shown in Fig. 17 (Ref 45) for a broken single fibre embedded in a block of matrix. Each fringe or contour represents the loci of points with an equal value of shear stress. Since the stresses are not constant through the thickness of the specimen, the observed photoelastic shear stress values are actually an average Table 2. Property E1 (GPa) E2 (GPa) 12 G23 (GPa) oq (10-6 C-1) Fibre and matrix material properties 7 Epoxy 3.8 3.8 0.35 1.4 68 AS-4 241 21 0.25 8.3 -0.11
[C2+~t2Cl(4.75-l)e-4"75~]%
(18)
Constants C2 and C1 are dependent on material properties, thermal strains and the applied far-field strain. Numerical results are normalized by o o, which represents the far-field fibre stress in the absence of expansional strains. In particular: o0 = C3% The constants C1, (72 and (73 are given by: (19)
el=
Elf
1-
elf ~)
. + % /
4KfGmVl2f {
(Kf+ Gin)
(Vl2f'-Ym)
+ [(lWVm)Em--E2f-Vl2t'Elf]} Co 2gfGm C2= (gf+Gm) (Vl2f"Vm)-~ [(i
(20)
"4-~m)Em--EZfIf!}12t~Eo -V
(21)
12
COMPOSITES . JANUARY 1992
x/I
1.0
x/I c
-2 0.8
AT=-75oc,c =1%
o
0.6 o u,_ o 0.4
x
-4
-6
%
xll
c~
t_
0.2
-8
-10
I
0 0.25 0.50
I
0.75
I
I .00
I I
1.25 1.50
xll
-12
0 0.25 0.50 0.75 1.00
I
1.25 1.50
Fig. 14 Distribution of axial stress along fibre fragment

length 7 (AS-4/Epon 828, A T=0)
0.0
x/I
Fig. 16 Distribution of radial stress along fibre fragment length7 (AS-4/Epon 828)
-0.5
o
c~
-1.0
~-
-1.5
-2.0
-2.5
a
I 0.25 1 0.50 I 0.75
c
-3.0
x/I
I 1.00
I 1.25 1.50
Fig. 15 Distribution of shear stress along fibre fragment length 7 (AS-4/Epon 828, A T=0) effect and had to be corrected as reported elsewhere 45"46. The shear stress distribution along the embedded length of the fibre obtained from both the photoelastic analysis and finite element analysis are shown in Figs 18 and 19 respectively. The shear and radial stresses are normalized by dividing by the far-field matrix tensile stress and the position normalized by dividing by the diameter of the fibre. It is easy to observe that the maximum shear stress is located at some distance from the broken end of the fibre and it decreases at points located further away from the break. The location of the peak stress occurs at approximately 0.2 fibre diameters away from the end of the broken fibre. This high peak shear stress is responsible for plastic deformations of the matrix, especially in strongly bonded systems, and for interfacial crack growth in weakly bonded systems. In an actual fibre fragmentation experiment only a very intense light area is observed around the fibre/matrix interface instead of fringes as depicted in Fig. 17. Normal practice is to use a white light source with a regular polarized optical microscope. This results in coloured isochromatic
x "C_
Fig. 17 Isochromatic fringe pattern for a single broken fibre obtained from: (a) photoelastic analysis; (b) finite element analysis. Insert shows area studied
fringes in a very narrow area too small to be resolved by commonly available photographic techniques, thus resulting in a white light image.
COMPOSITES . JANUARY
1992
13
3.5
4.0
3.0
3.5
3.0
2.5
2.5
2.0
o >~
o >,
2.0
1.5
1.5
1.0
1.0
0.5
0.5
I
2
I
4
I
6
x/d
I 0
Fig. 18 Interfacial shear stress distribution along the fibre/ matrix interface along a single broken fibre obntained from photoelasticity analysis
I 1
I 2
x/d
Fig. 19 Interfacial shear stress distribution along the fibre/ matrix interface along a single broken fibre obtained from finite element analysis
Fig. 20 shows transverse (radial) normal stresses along the interface for the single broken fibre problem. Near the end of the broken fibre, it is seen that this stress component is tensile, drops rapidly to zero at approximately one-fifth of a fibre diameter, and changes to compressive in a damped oscillatory fashion. This indicates that the tensile transverse stress at the crack tip could contribute to debonding at the interface, but not along the region where it changes to compression. It should be pointed out that thermal residual stresses resulting from the shrinkage of the matrix were not included in this finite element analysis. It has been shown by Whitney and Drzal 7 that the radial thermal stresses resulting from curing of the matrix remain compressive along the fibre and should cancel out some of the tensile stresses at the crack tip due to the mechanical loading, perhaps making them negligible. However, at other points along the fibre fragment the thermal and mechanical stresses will couple, increasing the frictional component of the interracial stresses. During fragmentation of a fibre embedded in a block of matrix, a penny-shaped crack may result. Depending on the amount of energy released upon fibre failure and the level of adhesion between fibre and matrix, this microcrack may or may not propagate into the matrix and/or the interphase. Thus the stress and strain
distribution will depend on the mechanical properties of the fibre, interphase and matrix and the extent of damage in the neighbourhood of the fibre break. As pointed out by Ko 76, when the interfacial shear strength or the matrix shear strength is low, the open mode deformation of the penny-shaped crack could induce interfacial or near interfacial shear failure, deflecting the original crack propagation direction to that of the interface. Thus, any further interfacial debonding should be caused by interfacial shear stresses rather than radial stresses.
Statistical characterization of the critical fibre fragment length distribution

Most high-performance engineering fibres have strengths that vary considerably along their length because of flaws introduced through handling or manufacturing, or because of intrinsic anomalies of the material 77"78. These randomly spaced flaws or point defects introduce a slight variation in strength which, depending on its value, may or may not prevent the fibre from fracturing once the average of is reached. When the built-up stress in the fibre approaches the true value of of, the fibre will instantaneously and repeatedly break into shorter and shorter pieces until
14
0"61
L Y ?:i
0.4
~ i / / : i/i !?
0.2
Bascom and Jensen 79 used an approach similar to that of Drzal and co-workers 13. Wimolkiatisak and Bell s() found that the fragmentation length data fitted both the Gaussian and Weibull distributions equally well. Fraser e t al. sl developed a computer model to simulate the stochastic fracture process and, together with the shear-lag analysis, described the shear transmission across the interface. Netravali et al. 77 and Henstenburg and Phoenix 7s used a Monte Carlo simulation of a Poisson/Weibull model for the fibre strength and flaw occurrence to calculate an effective interfacial shear strength xi using the relationship: l:i=K2f ( d )
(26)
0.0
o
O
L.
-0.2
where K is a correction factor to be determined from the Weibull/Poisson model and l is the mean fibre fragment length. They proposed that a value of 0.889 is an appropriate correction factor (K) for brittle fibres. The experimental determination of the strength of individual fibres at very short lengths is, however, very difficult, and most analyses extrapolate mean strength and strength distribution data obtained from longer gauge lengths. Asloum e t al. 8~ studied the dependence of the strength of high-strength carbon fibres on gauge length by means of the Weibull model. They showed that the mathematical form of the estimator chosen and the sample size, when larger than about 20, do not influence the results of the analysis. Also, they found that neither the three-parameter nor the two-parameter Weibull distribution is appropriate for describing the critical length dependence on gauge length of the fibre during testing. They also showed that a linear logarithmic dependence of strength on gauge length is a simple, yet the most accurate method for extrapolating the fibre tensile strength at short lengths. There are, as can be seen, different statistical models that have been used to describe the distribution of the fibre fragment lengths on the single-fibre fragmentation test. Despite slight differences between the statistical treatments of the experimental data, they should prove to be more useful if they are combined with a more accurate theoretical model of the mechanical events that occur in this technique.
-0.4
-0.6
-0.8 I 0
I 1
I 2
x/d
q 3
LI
Fig. 20 Interfacial radial stress distribution along the fibre/ matrix interface along a single broken fibre obtained from finite
element analysis the remaining fragments are not long enough for the linear stress to build up from either end to exceed the fibre strength anywhere. This final length is usually referred to as the critical length, lc. Any fragment with a length slightly exceeding lc will break in two, yielding, at the conclusion of the experiment, a random distribution of fragment lengths between lc/2 and lc. Drzal e t al. 1.13recognized the random nature of this problem and used a two-parameter Weibull distribution to characterize the distribution of fibre fragment length. Then, using the arithmetic mean fragment length (i.e., the original unbroken length divided by the number of fragments) and an average value for of at the critical length (Ic) extrapolated from simple tension tests, they obtained expressions for the mean interfacial shear strength:
=
Measurement of the critical fibre fragment length

The fibre fragment lengths have to be measured in the single-fibre fragmentation test specimen. Although a large statistical sampling of the interface occurs with the fragmentation test (for some matrix combinations, the number of breaks is of the order of 50 to 100 in a gauge length of 20 mm), this process can be tedious and time consuming because each test coupon must be analysed individually. Traditionally, the fibre failure positions have been measured optically using a microscope equipped with a calibrated filar eyepiece, leaving room for errors due to the inherent limitations of light microscopes. The optical method also requires that the matrix be transparent in order to experimentally interpret the failure mode of the fibre/matrix interface.
o, ( 1 )
r 1-
(24)
02 Var(x)=4--~ { r
(1-2 a)
_rz
(1-
1)}
(25)
where 13and 0are the maximum likelihood estimates of the scale and shape parameters, respectively, and r is the gamma function.
COMPOSITES
. JANUARY
1992
15
An acoustic emission technique (AE) has been developed recently for the measurement of fibre fragment length distributionss3. This technique is based on the fact that the speed of wave propagation is a function of the specimen material itself. It may also be influenced, however, by geometrical parameters, dominant frequency of the emitted ultrasonic signal and, more important, the deformation of the material. A simple algorithm incorporating the average wave speed in the epoxy (or any other matrix), the distance between the two receiving transducers, the offset distance between one specific receiving transducer and the fixed grip, time intervals and the corresponding strains was used to obtain the location of the fibre breaks, the fibre fragment lengths and fibre aspect ratios. A comparison of aspect ratios measured by optical and acoustic emission methods for glass fibres in two different epoxy blends was made (Fig. 21). Some discrepancies were obtained between the acoustic emission and optical techniques for low aspect ratios in the brittle epoxy blend. This may have been because one of the AE probes used had a diameter of 1 mm. Better agreement was obtained for the interracial shear strength for the flexible resin blend using both acoustical and optical techniques. One advantage of the acoustic measurement technique is that it does not require a transparent matrix and therefore can be used in matrices such as metal and many polymeric matrices. The sensitivity of this A E method in determining position, however, limits its use to large diameter fibres. An alternative technique to measure the fibre fragment length has been developed by Waterbury and Drzal s4. It uses a special software package called 'Fibertrack',
In (aspect ratio} 2 3 4
I I I I Jl
together with a computer-interfaced translation stage. While the coupon is translated at a constant velocity, the operator presses a 'mouse' button as each break passes a set of cross hairs. The time intervals between breaks are stored in RAM memory and converted into displaced distances by entering the total distance travelled as noted from the motion controller readout. These distances are saved to a diskette and combined with fibre strength and diameter data to produce Weibull distributions and gamma function calculations to find the interfacial shear strength. This technique thereby makes use of the human ability to discriminate events and to identify breakpoints visually while freeing the operator from much of the drudgery associated with manual operation of the test. The entire process of test coupon mounting, loading, fibre fragment length measurement, data storage and interracial shear calculation requires approximately 6 min per fibre, which represents an improvement of one order of magnitude over less automated methods. Photoelastic evaluation of the interface
When undeformed, some common polymeric matrices
can be considered optically isotropic. However, when subjected to stresses--whether due to externally applied loads or thermally induced stresses from differential shrinkage during sample cool-down from cure temperature--the material becomes optically anisotropic (birefringent). If the resin is sufficiently s transparent, it can be studied with polarized light 1-4's- . Drzal et al. t-n.13 observed qualitative differences in the
In (aspect ratio}
1
I
5
I
45
0.999 I 0.99
O.
0.99910.99
0.9
I
O Optical [3 Acoustic
O Optical [] Acoustic
O.
0.5
L~
13
-- 2
.~
m
o.1 D
[]
~ c
._ ~ 0.1-
-2
u~ r
I
T
_
~
o_ O
o
13
-4 []
0.01
--4 o 0.01 []
-- -6
o.ooi[ ~ I
~ 2
~ 5
~ 10
~ ~ 20
~ 50
~ - 6 100 200
0.001 500
/I 5
I 10
I 20
I 50
100
200
500
Aspect ratio
Aspect ratio
Fig. 21 Comparison of fibre fragment lengths measured using acoustic emission and optical techniques
16
COMPOSITES. J A N U A R Y 1992
stress pattern resulting from interface changes when working on graphite fibres and epoxy matrices. Normal chromatic cycles were observed in the epoxy at low strains, but they disappeared at higher load levels. Isochromatic fringes were not observed under high strain conditions present at the fibre fragment ends. Instead, a light, ellipsoidal region was observed. Fig. 22 shows a series of photoelastic stress patterns with increasing strain. The photoelastic data were generated with untreated fibres (AU) in an epoxy matrix. At low strains after a fibre break, extensive resin birefringence can be seen around the fibre ends. With increasing strain, this birefringence activity rapidly extends down the fragment away from the break. Dynamic observation of this process suggests that the stresses proceed along the fibre fragment by a stick-slip mechanism. That is, the stresses build up and then appear to release and move ahead an incremental amount, repetitively. For this particular fibre combination, each fibre fragment fails interfacially at low strain levels. With increasing load, the fibre fragments interact with the matrix only through weak interfacial frictional forces, resulting in a very low interfacial shear strength.
Strain
Surface-treated fibres (AS), which are mechanically identical to the AU fibres, behave in a completely different manner to the untreated fibres under polarized light. Fig. 23 shows that the stresses build up at the end of the fibre at the fibre break. However, at higher strain levels, a narrow, very intense region of photoelastic activity remains around the fibre, while the initial ellipsoidal region moves away from the fibre ends toward the centre of the fibre fragment. The interfacial shear strength that was measured from this sysem was more than three times greater than that obtained for the AU fibre, indicating a higher degree of fibre/matrix interaction. Subsequent examination of ultramicrotomed sections of each fibre/matrix interphase and comparison with the photoelastic photographs suggest that the narrow intense area that remains around the fibre, especially at the tip of the broken fibre, is a region where an interfacial crack has passed or where permanent shear deformations have been induced. It is thus evident that photoelastic observations before, during, and after application of loads to a single-fibre composite coupon could help to elucidate the fibre/matrix interactions, as well as to judge the effect of surface chemistry and morphology on the properties of the fibre/matrix interface. Waterbury has noted that comparisons must be made at equal
Strain
~
1.5%
T.5%
2.5%
~ 2.0%
3.0%
3.0%
4.0%
3.5%
5.0%
4.5%
6.0%
6.0%
1 O0 IJrn
100 urn
Fig. 22 Photoelastic patterns obtained for AU-4 carbon fibres in an epoxy matrix (Epon 828 cured with mPDA)
Fig. 23 Photoelastic patterns obtained for AS-4 carbon fibres in an epoxy matrix (Epon 828 cured with mPDA)
COMPOSITES
. JANUARY
1992
17
levels of strain to avoid artefacts resulting from differences in the birefringent patterns s6.

Thermoset test coupons for the single-fibre fragmentation test can be fabricated by a casting method with the aid of a silicone room temperature vulcanizing (RTV) 664 eight-cavity mould. Standard ASTM 50.8 mm long dogbone specimen cavities with a 3.175 mm wide by 1.587 mm thick by 25.40 mm long gauge section are moulded into a 76.20 x 203.20 x 12.70 mm silicone piece. Sprue slots are moulded in the centre of each dogbone to a depth of 0.7938 mm and through the end of the silicone piece. The procedure is as follows. 1) Single fibres approximately 150 mm in length are selected by hand from a fibre bundle. Single filaments are carefully separated from the fibre tow without touching the fibres, except at the ends. Once selected, a filament is mounted in the mould and held in place with a small amount of rubber cement at the end of the sprue. The rubber cement is not in contact with the cavity which contains the grip sections nor the gauge length section in the mould. 2) The rubber cement is allowed to dry, and the resin is added with the aid of a disposable pipette. The long, narrow tip should be removed so that the resin can readily enter and exit the pipette chamber. Air bubbles are avoided by first degassing the silicone mould and the resin in a vacuum chamber before filling the mould cavities. 3) The assembly is transferred to an oven where the curing cycle is completed. After cool-down to room temperature, the mould can be curled away from the specimens parallel to the fibre to prevent fibre damage. The test samples can then be stored in a desiccator until ready for analysis. 4) Prior to testing, the coupons should be inspected for defects, and any defective ones should be discarded. The single-fibre coupons used in this study were prepared following the same curing procedure as for the microdrop preparation. 5) The specimens are tested in uniaxial tension using a microstraining machine capable of applying enough load to the tensile coupon (Fig. 24). The coupon is
fitted to the microscope stage so that the X - Y stage controls manipulate the jig position. This allows the operator to assess the fibre fracture process along the entire gauge length of the coupon. A transmitted light polarizing microscope should be configured such that there is one polarizer below and one above the test coupon. Since the embedded fibre is located in the centre of the polymeric coupon and therefore is difficult to observe at high magnification with standard objectives, the microscope should be equipped with a long working distance 20x or greater objective lens. The fibre diameter is measured using a calibrated filar eyepiece. A method to fabricate single fibre fragmentation specimens using a thermoplastic material and glass or carbon fibre reinforcement is described below. 1) On a flat surface, place a 300 x 300 x 0.2 mm plate of aluminium. Place a sheet of Kaptan film of the same dimensions over the aluminium plate. 2) Put on top, a 200 x 200 x 0.9 mm sheet of iron-stabilized silicone rubber (from Exotic Rubbers) with a 100 x 100 mm square removed from the centre. Into the hole of the rubber gasket place one 100 x 100 mm sheet of polycarbonate. 3) Lay fibres across the rubber gasket and the thermoplastic film. The fibres are held in place by static electricity forces between the fibres and the thermoplastic. The spacing between the fibres should be approximately 10 mm. 4) Once all the fibres are in place, cut them with a scalpel into 25 mm long pieces. This helps in keeping the fibres straight during compression of the laminate. 5) Place the thermocouple over one corner of the thermoplastic. It can be held in place by taping it to the gasket material using high temperature tape. 6) Complete the assembly with another sheet of thermoplastic, Kaptan film and aluminium plate. 7) Enclose this assembly in a bag of high temperature material (include breather cloth) that is formed by attaching a vacuum hose to the bag, sealing the edges and ensuring a perfect seal. 8) Pull a vacuum, heat the press platens to the desired temperature and place the assembly in between them so that good thermal contact is established but no pressure. The sample temperature must be monitored and maintained at this temperature to allow the thermoplastic to dry. 9) Increase the temperature of the platens to above the Tm of the material. Once at temperature apply a compressive force of 33 400 N to the assembly. Hold the sample at this pressure for i h, and then quench it to room temperature by quickly cooling the press platens (20 min approximately). 10) Examine the laminate and select fibres which remain straight. Dogbone-type specimens of selected fibres (with same dimensions as those specified for thermoset samples) can now be cut from the sheets.
Fig. 24 Single-fibre fragmentation technique apparatus: 1--displacement gauge; 2--single fibre specimen; 3--grips; 4--loading screw
18
MICRODEBONDIMICRO-INDENTA TION TECHNIQUE

Mandell and co-workers ~'s7-9 first proposed an alternative technique to measure the interfacial shear strength. Single fibres perpendicular to a cut and polished surface of a regular, high fibre volume fraction composite are compressively loaded to produce debonding and/or fibre slippage s7 as shown in Fig. 25. In contrast to conventional methods, which use a model system to provide information on fibre/matrix adhesion, this micro-indentation technique is an in situ interface test for real composites and has the advantage of reflecting actual processing conditions. It can allow determination of the interface strength due to fatigue or environmental exposure, or possibl~ monitor interface properties of parts in service ss. An analytical closed-form solution for the stress distribution around the indented fibre is still not available. The calculation of an average shear stress value relies on a finite element code 14.
Theoretical c o n s i d e r a t i o n s
01
Specimensurface
0.2
I
t~
0.4
0.6
The micro-indentation test is run on individual selected fibres on a polished cross-section of a high fibre volume fraction composite. An individual fibre is surrounded by neighbouring fibres located at various distances, and distributed in a variety of arrangements, which range from a hexagonal array to random and non-uniform configurations. The diameter of the test fibre and the distance to the nearest neighbour fibre are recorded for each test on a micrograph, and a simplified axisymmetric finite element model (FEM) is used. This model includes the fibre, surrounding matrix, and average composite properties beyond the matrix 88. It was shown that the maximum shear stress along the interface is insensitive to probe stiffness as long as the contact area does not approach the interface. Fig. 26 shows results for a case in which the fibre and matrix are considered to have the same mechanical properties. The finite element results are compared with those obtained from an analytical solution of the Hertz contact problem of a point load on a half-space with imaginary boundaries. G o o d agreement was found
0.8
1.0 0.1
0 Tension
0.1
0.2 Compression
0.3
Normalized stress Fig. 26 Interfacial stress components along an imaginary interface normal to the free surface in a homogeneous isotropic material: comparison of Hertzian point-load solution (--) with finite element results ( - - - ) . D is fibre diameter 9
between the FEM and the analytical solution despite slight differences in loading conditions. Figs 27 and 28 show results for Nicalon (SIC) fibres in an aluminosilicate glass matrix and for H M U carbon fibres in a borosilicate glass matrix system, respectively. The stress distribution for the Nicalon fibres is very similar to that for the isotropic homogeneous case because of the low Ef/E m ratio. It can also be noticed that the anisotropy of the carbon fibres affects the stress distribution by spreading out the shear stress transfer along the length of the fibre. Fig. 29 shows interfacial shear stresses for a carbon fibre/epoxy system. The low modulus of the epoxy matrix produces a similar effect on the stress distribution as the isotropic fibres. Calculations of interfacial shear strength assume either a maximum interfacial shear stress criterion or a maximum radial tensile stress criterion, ignoring other stress components and residual stresses in both cases. The interfacial shear strength (xi) is calculated from:
1;i = Ofd~--'Z~ ) \ of / F E M
T = f (Pdebonding)
/Xmax \
(27)
Fig. 25 Schematic diagram ofthe micro-indentation method whereby a single fibre is compressed for measuring interfacial shear strength
where Ofd is the average compressive stress applied to the fibre end at debonding and (Tmax/Of)FEM is the ratio
19
0.0
Specimen surface
F
/ d rr
Specimen surface
\
O 0
0.2
0.2
""%e
I
0.4
Trz 0.41Q
0.6
f, drr
I
0.8
/ OZZ
0.61-
0.8 1.0 I / ~ /rf I I
0.05
0.05
0.10
0.15
0.20
0.25
0.30
Tension
Compression
~I .0 II I
Normalized stresses
Fig. 27 Interracial stress distribution for Nicalon (SIC) fibres in an aluminosilicate glass matrix obtained using the microindentation technique9
0.02
0.04 0.06 0.08
0.10
Compression Normalized stresses
,,-
Specimen s u r f a c e
Fig. 29 Interfacial stress d i s t r i b u t i o n f o r a c a r b o n f i b r e / e p o x y m a t r i x s y s t e m 9
0.2
~r r ~ / ' ~ I\
of the interracial tensile strength ori for tensile radial stress at the surface is given by: /o#max~
I 0.4
I/ 1
0.6
//
O.
where (Or/max/(~f)FE M is the ratio of the maximum radial tensile stress at the surface to the applied fibre pressure resulting from the finite element analysis. In all cases, the FEM results are calculated for the ratio of matrix layer thickness to fibre diameter of 0.40. It is evident that further refinements of the finite element model such as the free-surface problem and the effect of thermal elastic residual stresses should be included because they should be very significant near the free surface; and failure criteria for fibre/matrix debonding, etc, should be included to obtain a more accurate distribution of stress at the fibre/matrix interface. To avoid any cyclic loading on the specimen during visual detection of fibre debonding, Netravali e t al. 91 developed a test procedure to monitor the load and depth of indentation continuously as a characteristic change in the load/depth curve. They used a slightly modified micro-indentation technique where the sample is sectioned to very thin dimensions to determine the interracial shear strength between E-glass fibres and an epoxy matrix. In addition, acoustic
o00 \
/[
I
0.05 0
"~O2
1.0 I 0.15 0.10 t
f,
0.05
I 0.10 0.15
I I 0.20 0.25 0.30 ~,
Tension
Compression Normalized s t r e s s e s
Fig. 28 Interfacial stress d i s t r i b u t i o n f o r H M U c a r b o n f i b r e s in b o r o s i l i c a t e glass m a t r i x o b t a i n e d u s i n g the m i c r o - i n d e n t a t i o n t e c h n i q u e 9
of the maximum shear stress to applied stress. Fig. 30 shows normalized interfacial shear stress ('Cmax/Of)as a function of Gm/E f (matrix shear modulus/fibre axial elastic modulus) for a variety of materials. Calculations
20
COMPOSITES . JANUARY
1992
Im
,~ ,
--~
Matrix~]] Fibre ~ 0.30
|j',/ FEM model
! ! t ' ~z Hertzian, homogeneous

5
bends upon application of the indentation force, resulting in radial compression at the top and tension at the bottom. The resulting stress distribution is calculated from the elastic analysis of the theory of plates according to: 0r=(6~)(2~){~+ [~]ln(d ) }(31)
0.25
0.20
where o r is the radial stress at the top surface, d is the fibre diameter, t is the specimen thickness and p is the inside radius of the brass annulus on which the specimen is rigidly mounted.
"x 0.15
0.10
1

2 9~
0.05
0.1
0.2
0.3
FEM
0.4
0.5
0.6
(Gm/Ef)
(Grr/Ez) Hertzian
Fig. 30 Maximum normalized interfacial shear stress for various hypothetical material systems and trends from Hertzian solutions for transversely isotropic properties reference to an imaginary interface position. (O) FEM model, isotropic fibres; (@) Hertzian with Gzr = 14 MPa, Ez varying; (A) FEM with Ez = 276 GPa, matrix varying; (A) FEM with Gm= 28 GPa, carbon fibre Ez varying; (1)S-glass/epoxy; (2)E-glass/epoxy; (3) Nicalon/1723 glass; (4) Nicalon/BMAS; (5) Hertzian isotropic; (6) carbon/epoxy; (7) P100/aluminium; (8) P55/aluminium; (9) HMU/borosilicate9
emissions generated by various events are monitored and the rate of loading and specimen geometry can be adjusted to simulate different situations that can occur in actual service. Thin samples approximately 4 to 10 fibre diameters thick were used, and the assumption of constant shear stress along the fibres was made. The shear stress was calculated from the relation: ,%_ F ndt (29)
where F is the load value, d is the diameter of the fibre and t is the thickness of the specimen. Differential shrinkage between the epoxy and the fibre after elevated temperature curing generates a hydrostatic pressure P, which is given by:
e = (XrnemA T(1 ~t_ Vm ) (30)
where o% is the linear coefficient of thermal expansion of the matrix material, E m is the matrix Young's modulus, A T is the difference between the curing and room temperatures, and vm is Poisson's ratio of the matrix. It should be noticed that the linear coefficient of thermal expansion of the fibre is considered to be negligible and also that its elastic modulus is several times greater than that of the matrix. As a consequence, lateral expansion of the fibre at the point of application of the force was also considered negligible. Because of the thinness of the sample, the specimen
The microdebonding indentation system, described here 92, overcomes some deficiencies of the method described by Mandell and co-workers. This fully automated instrument, designed to be used outside the research environment93"94, is based on a Mitutoyo Finescope. A diamond-tipped stylus mounted on a collar fitted to the objective of the microscope is used to compress single fibres into the specimen. The probe has a 90 cone with a user-selected tip radius. Initiation of fibre debonding is sensed by a weighing mechanism from a Sartorius model L610. Other components include a precision-controlled motorized stage with three degrees of freedom (linear motion in three orthogonal axes). Klinger linear motion stages for translations in the x and y axes replace the usual stage. The fine focusing control of the microscope is controlled by a Klinger stepping motor with +0.04 ~m resolution. The stage controllers and balance read-out are interfaced to a Zenith Z-386/20 microcomputer. The specimen surface is monitored using a video camera mounted on the microscope. The specimens are prepared following standard metallographic techniques, assuring that the fibres of the composite are always perpendicular to the surface of the specimen holder. The force (fg) required to debond the fibre is input to an algorithm that calculates the interfacial shear strength as a function of the fibre diameter (d0, the shear modulus of the matrix (Gin), the axial tensile modulus of the fibre (El) and the distance from the tested fibre to the nearest adjacent fibre (Tm). The fibres to be indented must be in the neighbourhood of other fibres but no closer than 2 ~m and no further than half a fibre diameter. Once the fibres are selected, their coordinates, fibre diameter and nearest neighbour distance are recorded. Fibre/matrix properties are also entered for analysis. The program directs the controller to move the fibre end selected through an offset to place the fibre below the indenter tip. The sample is moved to within 4 ~tm of the indenter, at this point the stage is slowed to the rate and step size selected by the user. Once the indenter contacts a fibre, a real-time plot of load versus displacement is obtained. Fibre debonding is visible as a dark shadow around the fibre. When the shadow appears around at least one-sixth of the fibre it is said to be debonded. Upon debonding, an interfacial shear strength is calculated and for a particular fibre/matrix combination an average value is obtained from individual fibres. The sample preparation procedure is as follows.
21
1) Square chips of composite laminate approximately 12.7 mm per side are cut using a water-cooled diamond saw. The squares are trimmed with a scalpel to remove burrs from the cutting process. 2) Tape is placed over one end of a cylindrical section of phenolic pipe 25.4 mm in diameter and the squares are held in place using spring holders. The squares are oriented with the fibre axis parallel to the pipe axis. The pipe section is filled with a low exotherm epoxy resin (9 parts DGEBA (Shell Epon 828) to i part diethylene triamine (Aldrich Chemical)). The filled samples are left to cure at room temperature overnight. 3) Specimens are polished using a Struers Abramin counter-rotating polisher. The polisher is set at a force of 50 N and a speed of 150 rpm and lubricated with water. Samples are polished using paper with grits 1000, 2400 and 4000 for 4, 5 and 6 min respectively. A final relief polishing step is performed using a Vibromet I polisher with 0.05 ~tm ~,-alumina suspended in water. This final step is carried out overnight and then the samples are thoroughly rinsed with distilled water, air dried, and placed in a desiccator until needed for testing.
shows the geometry of the specimen used in this test. Details of the test are given in ASTM D3518.
2. Iosipescu shear test specimen

A state of pure shear is achieved within the test section of the Iosipescu shear test specimen (Fig. 31) by applying two counteracting moments produced by two force couples 98. From simple statics considerations, the shearing force acting at the centre of the cross-section is equal to the applied force measured from the loading device. As stated by Iosipescu, there is no stress concentration at the sides of the notches of the specimen, because the normal stresses are parallel to the sides at that point of the specimen. Consequently, the shear stress is obtained by simply dividing the value of the applied force by the net cross-sectional area between the two notch tips.
3.
Short-beam shear test
Composite laminate techniques for shear characterization

Several methods have been developed to measure shear properties of composites laminates and some of them have been adopted as ASTM standards. The importance of these methods stems from the need to generate reliable design data especially when assessing the level of fibre/matrix adhesion with new sizing systems for the surface treatment of reinforcing fibres. A few of these methods have shown to be more promising than others because of their reliability and simplicity in sample preparation and data reduction. In this study single-fibre composite interfacial testing results will be compared with three most widely used laminate techniques. These techniques are first discussed briefly below.
The short-beam shear method is used to estimate interlaminar shear strength only. It is based on elementary beam theory and involves a three-point flexure specimen with a span-to-width ratio (L/h) chosen to produce interlaminar shear failure. The shear distribution along the thickness of the specimen is a parabolic function symmetrical about the neutral axis where it attains a maximum, and has a value of zero at the upper and lower surfaces. When used in conjunction with
Load direction
1. (+45)s tensile test

Based on expressions developed by Petit 95 for the shear stress/strain results of a uniaxial tensile test for a (_+45)
i
Pb/ ( a - - b } ~ P/2
Fig. 31 tests
~ibredirection
[-+45] s
tension shear
test
Pb/ (a-b)
h
symmetrical laminate using plate theory, Rosen 96 showed that the expressions for the in-plane shear stress/strain curve could be obtained from the longitudinal and transverse stress/strain curves as: 1712= T
(~x
losipescu shear test

J
I'b/ { ~r.b }
Y12 = (Cx-- Ey) where Ox is the applied tensile stress and ~x and Cyare the longitudinal and transverse strains. Caution should be exercised regarding the interpretation of the ultimate stress and strain results from this test because the laminate is under a combined state of stress rather than pure shear 97 and the presence of normal stress components would be expected to have a deleterious effect on ultimate shear strength. Fig. 31
Fibre direction
P/2
Short beam interlarninar shear test
Schematic representation of laminate c o m p o s i t e shear
22
COMPOSITES . J A N U A R Y 1992
laminated materials, the interlaminar shear stress will be parabolic within each layer, but a discontinuity in slope will occur at the ply interfaces. As a result, the maximum value of the shear stress will not necessarily occur at the centre of the beam. Thus, caution should be exercised when interpreting the results of composite materials that cannot be treated as homogeneous 97. The interlaminar shear stress is given by: Zm= 0.75 . W W A where " 7 is the maximum shear stress, W is the applied 1m load and A is the cross-section of the specimen. Details of this test are documented in ASTM D 2344. The geometry of this specimen is given in Fig. 31. COMPARISON OF TECHNIQUES In this study to compare the various methods, carbon/ epoxy and S-glass/polycarbonate composite systems were used. The 'A' type carbon fibres were polyacrylonitrile-based graphite, designated AU-4 and AS-4 (Hercules, Inc). These fibres differ only in their surface treatment. The 'AU' fibres are 'as-received', that is, removed from heat treatment ovens without any further processing. 'AS' fibres are surface-treated with a commercial electrochemical oxidation step which optimizes the adhesion to epoxy matrices. An amine-epoxy matrix system was selected for this study. A difunctional epoxy, DGEBA (Epon 828, Shell Chemical Co), was processed at stoichiometric conditions (14.5 phr mPDA, Aldrich Chemical Co) for both single fibre and composite test coupons for 2 h at 75C and 2 h at 125C. Values of interfacial shear strength measured with the single-fibre fragmentation technique were used as a reference to compare the results from the single-fibre techniques. Table 3 shows a summary of test results obtained from the different single-fibre and laminate techniques for the carbon fibre/epoxy matrix composite system. Comparing the interfacial shear strength values obtained from the first and second single-fibre techniques, it can be seen that the microdrop technique yields lower values. This difference can be explained if it is remembered that the bulk properties of the microdrops for this epoxy system are slightly lower than those of the matrix surrounding the single fibre in the
fragmentation technique 99 because of loss of curing agent in the small drops having large surface-to-volume ratios. This diffusion problem has not been reported by other investigators in their work on carbon fibre/epoxy resin systems n'52, although a different curing agent was used. There are also differences in the events that take place during failure for both tests 45. In the microdrop technique, the total fracture energy is contributed by the strain energy stored in the free and embedded length of the fibre and in the surrounding matrix. In the singlefibre fragmentation method, the total energy is contributed by the embedded fibre and the surrounding matrix. The micro-indentation technique yields slightly higher values than the single-fibre fragmentation technique. However, some further refinement is necessary in order to completely assess the effect of other factors such as debonding criteria, free surface, residual stresses, effect of neighbouring fibres, and possible damage to the surface during the polishing of the specimen. When probing the interphase with the microindentation technique, a failure criterion has to be defined in terms of load drop during testing. This means that, as soon as the force applied to the fibre tip drops a preset amount, the test is stopped automatically and the interfacial shear strength calculated. It has been shown from photoelastic observations of the fragmentation test that for fibres which exhibit low adhesion like the AU-4 fibres, an interfacial crack propagates rapidly at low load levels. For the case of AS-4 fibres, however, a higher load is required to produce debonding and considerable damage to either the fibre or matrix could also be produced in some cases. Nevertheless, the computer program will calculate an interface strength value as soon as the preset load drop is reached, regardless of the extent of debonding of the interface. When the adhesion level between fibre and matrix is increased, the failure criterion should be changed accordingly, otherwise this limitation in the data reduction program could lead to errors in the numerical results obtained. It is interesting to observe that the interfacial shear strength almost doubled its value with the surface treatment (AS-4) as measured with the single-fibre fragmentation method, but the micro-indentation technique reflects only a 17% increase due to the
Table 3.
Test
Summary of interfacial shear strength values

AU-4/828, mPDA (MPa) 37.20 23.44 55.54 37.16 54.98 47.52 AS-4/828, mPDA (MPa) 68.30 50.30 71.53 72.19 95,28 84,04 Ratio AS-4/AU-4 1.8 2.1 1.3 1.9 1.7 1.8
Single-fibre fragmentation test Microbond test Micro-indentation technique (+45)3s tension test Iosipescu shear test Short-beam shear test
23
surface treatment. This is an indication that each technique reacts differently to the conditions which prevail at the interphase. The results obtained from the single-fibre techniques were also compared with the three composite laminate techniques for shear characterization, namely, the short-beam shear test, the Iosipescu test and the off-axis [+45] tensile test. These results are also given in Table 3. The in-plane shear strength determined by the +45 tension test almost doubled from AU-4 to AS-4, and the Iosipescu shear test shows that the interfacial shear strength of the AS-4 composite is 1.75 times that of the AU-4 composite. It has been shown by Madhukar and Drzal 6 that in the case of poor adhesion between fibre and matrix, the +45 tension test will yield higher values of interfacial shear strength due to rigid body rotation of the individual plies caused by a 'scissoring' effect which results in lower strain readings during the experiment. There exists good agreement between the interfacial shear strength measured by the single-fibre fragmentation test and the interlaminar shear strength test measured by the short-beam shear test, since both measure the same component of shear stress (x13) despite the fact that both techniques are based on different equations and assumptions.
furthermore, it has also been shown by Rao et al. 99 that the mechanical properties of the microdrop may vary with size because of variations of concentration of the curing agent; and for a given fibre/matrix combination, a relatively large scatter in the test data is obtained from the microdrop test. Such wide distribution of shear strengths have been attributed mainly to testing parameters such as position of the microdrop in the loading fixture, droplet gripping, faulty measurement of fibre diameters, and so on '1. In addition, variations in the chemical, physical or morphological nature of the fibre along its length will affect the results of interfacial shear strength measurements, which only consider very small sections35"55. Single-fibre fragmentation technique The advantages can be summarized as follows: this technique yields a large amount of information for statistical sampling; as mentioned before, in the case of transparent matrices, the failure process can be observed under polarized light; this technique replicates the events in situ in the composite; and fewer parameters are involved in its characterization. The shortcomings can be summarized as follows: the matrix must have a strain limit at least three times greater than that of the fibre; the matrix should have sufficient toughness to avoid fibre fracture-induced failure; the fibre strength should be known at the critical length; transverse normal stresses could be higher due to Poisson's effects, resulting in a higher interfacial shear stress; the state of stress at the fibre break is highly complicated by the presence of the penny-shaped crack which could affect the failure mode of the interface, thus affecting the obtained values of interface strength; and despite the sophisticated statistical techniques used to characterize the fibre fragment length distribution, the shear strength is calculated using an oversimplified representation of the state of stress occurring at the interphase.
ADVANTAGES AND LIMITATIONS Microbond technique

The major advantages of the microbond techniques can be summarized as follows: the value of force at the moment of debonding can be measured; and this technique can be used for almost any fibre/ matrix combination. On the other hand, there are serious inherent limitations to the microbond technique: the debonding force is a function of the embedded length. When using very fine reinforcing fibres with diameters ranging from 5 to 50 ~xm,the maximum embedded length is of the range 0.05 to 1.0 mm. Longer embedded lengths cause fibre fracture; the meniscus that is formed on the fibre by the resin makes the measurement of the embedded length somewhat inaccurate; the small microdrop size makes the failure process difficult to observe; most important, the state of stress in the droplet can vary both with size and with variations in the location of points of contact between the blades and the microdrop; the presence of a meniscus has a large effect on the interfacial stresses which oscillate along the embedded fibre, thus making the practice of calculating an average shear strength value questionable. Also, premature failure of the meniscus region due to tensile stresses could lead to premature microdrop debonding off the fibre;
Micro-indentation/microdebonding
The advantages are:
technique
allows in situ measurement of debonding force; allows probing of the interface in the 'real' environment; it yields multiple data points; and
24
d a t a c o l l e c t i o n is fast a n d a u t o m a t e d . T h e d i s a d v a n t a g e s are: t h e f a i l u r e m o d e o r locus o f failure c a n n o t b e observed; \~ t h e r e exists t h e p o s s i b i l i t y o f i n d u c i n g a r t e f a c t s by the surface p r e p a r a t i o n p r o c e d u r e ; t h e a s s u m p t i o n s m a d e to c a l c u l a t e t h e i n t e r f a c i a l s h e a r stress m a y n o t b e valid; a n d crushing o f fibres is o b s e r v e d v e r y f r e q u e n t l y , limiting t h e v a r i e t y o f fibres to b e t e s t e d .
CONCLUSIONS
T h e single-fibre m o d e l s for i n t e r f a c i a l s h e a r s t r e n g t h m e a s u r e m e n t yield d i f f e r e n t levels o f sensitivity to c h a n g e s in f i b r e / m a t r i x a d h e s i o n a n d i n t e r f a c i a l failure m o d e . A l l i n d i c a t e d h i g h e r a d h e s i o n for t h e surfacet r e a t e d fibre o v e r t h e u n t r e a t e d fibre. F r o m the c o m p a r i s o n o f the s i n g l e - f i b r e m e t h o d s , t h e single-fibre f r a g m e n t a t i o n t e c h n i q u e for t h e m e a s u r e m e n t of t h e i n t e r f a c i a l s h e a r s t r e n g t h s e e m s to be m o r e sensitive to c h a n g e s in f i b r e / m a t r i x a d h e s i o n d u e to t h e effect of surface t r e a t m e n t o r finish if t h e c o m p o s i t e p r o p e r t i e s a r e to b e p r e d i c t e d . A l t h o u g h t h e single-fibre tests p r o d u c e d v a l u e s o f i n t e r f a c i a l s h e a r s t r e n g t h t h a t a r e v e r y s i m i l a r to c o m p o s i t e v a l u e s , r e l i a n c e on s i n g l e - f i b r e m e a s u r e m e n t s s h o u l d b e a v o i d e d until f u r t h e r r e s e a r c h is c o m p l e t e d to p r e d i c t c o m p o s i t e p r o p e r t i e s . It is clear t h a t n o n e o f t h e s e t e c h n i q u e s offers a c o m p l e t e a n d u n a m b i g u o u s m e t h o d for m e a s u r i n g t h e i n t e r f a c i a l s h e a r s t r e n g t h b e t w e e n fibre a n d m a t r i x . F o r a p r o p e r i n t e r p r e t a t i o n o f test results, t h e a d h e s i o n m e t h o d s e l e c t e d s h o u l d p r o v i d e a c l e a r i d e a o f t h e level of a d h e s i o n a n d failure m o d e b e t w e e n fibre a n d m a t r i x .
ACKNOWLEDGEMENT
T h e m a j o r p o r t i o n o f the r e s e a r c h r e p o r t e d in this p a p e r was s u p p o r t e d by t h e N a t i o n a l I n s t i t u t e of Standards and Technology under NIST Cooperative Agreement No 70NANB9H0952. The authors are a p p r e c i a t i v e of t h e a d d i t i o n a l s u p p o r t p r o v i d e d b y t h e Office o f N a v a l R e s e a r c h , T h e A l c o a C o m p a n y , D o w Chemical Company, and the State of Michigan Research Excellence Fund.
REFERENCES
1 Drzal, L.T., Rich, M.J. and Lloyd, P.F. 'Adhesion of graphite fibers to epoxy matrices: I. The role of fiber surface treatment' J Adhesion 16 (1982) pp 1-30 2 Drzal, L.T., Rich, M.J., Koenig, M. and Lloyd, P.F. 'Adhesion of graphite fibers to epoxy matrices: II. The effect of fiber finish' JAdhesion 16 (1983) pp 133-152 3 Drzal, L.T., Rich, M.J. and Koenig, M. 'Adhesion of graphite fibers to epoxy matrices. III. The effect of hygrothermal exposure' J Adhesion 18 (1985) pp 49-72 4 Drzal, L.T. 'Composite interphase characterization' SAMPE Journal 19 (1983) 5 Drzal, L.T. 'Effect of graphite fiber-epoxy adhesion on composite fracture behavior' in Proc 2nd US~JapanASTM Conf on Composite Materials (American Society for Testing and Materials, 1985)
6 Madhukar, M. and Drzal, L.T. 'Fiber-matrix adhesion and its effect on composite mechanical properties. I. Inplane and interlaminar shear behavior of graphite/epoxy composites' to appear in J Composite Mater (1991) 7 Madhukar, M. and Drzal, L.T. 'Fiber-matrix adhesion and its effect on composite mechanical properties. II. Longitudinal (0) and transverse (90) tensile and flexure properties of graphite/ epoxy composites' to appear in J Composite Mater (1991) 8 Madhukar, M. and Drzai, L.T. "Fiber-matrix adhesion and its effect on composite mechanical properties. III. Longitudinal (0) compressive properties of graphite/epoxy composites' to appear in J Composite Mater ( 1991) 9 Madhukar, M. and Drzal, L.T. 'Fiber-matrix adhesion and its effect on composite mechanical properties. IV. Mode I and Mode II fracture toughness of graphite/epoxy composites' to appear in J Composite Mater ( 1991) 10 Broutman, L.J. 'Measurement of the fiber-polymer matrix interfacial strength' Interfaces in Composites, ASTM STP 452 (American Society for Testing and Materials, 1969) pp 27-41 11 Miller, B., Muri, P. and Rebenfeld, L. 'A microbond method for determination of the shear strength of a fiber/resin interface' Composites Sci and Techno128 (1987) pp 17-32 12 Kelly, A. and Tyson, W.R. 'Tensile properties of fiberreinforced metals: coppe/tungsten and copper/molybdenum' J Mech Phys Solids 13 (1965) pp 329-350 13 Drzal, L.T., Rich, M.J., Camping, J.D. and Park, W.J. 'Interfacial shear strength and failure mechanisms in graphite fiber composites' 35th Ann Tech Conf Reinforced Plastics~Composites lnst (The Society of the Plastics Industry. 1980) paper 20-C 14 Mandell, J.F., Chen, J.-H. and McGarry, F.J. 'A microdebonding test for in-situ fiber-matrix bond and moisture effects' Res Rep R80-1 (Department of Materials Science and Engineering, Massachusetts Institute of Technology, Feb 1980) 15 Outwater, J.O. and Murphy, M.C. 'The influences of environment and glass finishes on the fracture energy of glass-epoxy joints' in Proc24th Ann Tech Conf(The Society of the Plastics Industry, 1969) paper 16-D 16 Narkis, M., Chert, E.J.H. and Pipes, R.B. 'Review of methods for characterization of interfacial fiber-matrix interactions' Polym Composites 9 No 4 (Aug 1988) pp 245-251 17 Ko, Y.S., Forsman, W.C. and Dziemianowicz, T.S. 'Carbon fiber-reinforced composites: effect of fiber surface on polymer properties' Polym Engng and Sci 22 (Sept 1982) pp 805-814 18 Chua, P.S. ~Characterization of the interfacial adhesion using tan delta' SAMPE Quarterly 18 No 3 (April 1987) pp 10-15 19 Perret, P., Gerard, J.F. and Chabert, B. 'A new method to study the fiber-matrix interface in unidirectional composites: application for carbon fiber-epoxy composites' Polym Test 7 (1987) pp 405--418 20 Yuhas, D.E., Dolgin, B.P., Vorres, C.L., Nguyen, H. and Sehriver, A. 'Ultrasonic methods for characterization of interfacial adhesion in spectra composites' in Interfaces in Polymer, Ceramic and Metal Matrix Composites edited by H. Ishida (Elsevier, 1988) pp 595-609 21 Wu, W.L. "Thermal technique for determining the interface and/or interply strength in polymeric composites' Preprint (National Institute of Standards and Technology, Polymers Division 1989) 22 Greszczuk, L.B. 'Theoretical studies of the mechanics of the fiber-matrix interface in composites' in Interfaces in Composites, ASTM STP 452 (American Society for Testing and Materials, 1969) pp 42-58 23 Lawrence, P. 'Some theoretical considerations of fibre pull-out from an elastic matrix' J Mater Sci 7 (1972) pp 1-6 24 Takaku, A. and Arridge, R.G.C. 'The effect of interfacial radial and shear stress on fibre pull-out in composite materials' J Phys D: Appl Phys 6 (1973) pp 2038--2047 25 Gray, R.J. 'Analysis of the effect of embedded fibre length on the fibre debonding and pull-out from an elastic matrix' J Mater Sci 19 (1984) pp 861-870 26 Laws, V. "Micromechanical aspects of the fibre-cement bond' Composites 13 (April 1982) pp 145-151 27 Banbaji, J. "On a more generalized theory of the pull-out test from an elastic matrix. Part I, theoretical considerations' Composites Sci and Techno132 (1988) pp 183-193 28 Banbaji, J. 'On a more generalized theory of the pull-out test from an elastic matrix. Part II, application to polypropylenecement system' Composites Sci and Techno132 (1988) pp 195207
COMPOSITES
. JANUARY
1992
25
29 Hampe, A. 'Grundlegende Untersuchungen an faserverstarkten Polymeren' (BAM, Federal Institute for Materials Research and Testing, FRG, August 1987) 30 Chua, P.S. and Piggott, M.R. 'The glass fibre-polymer interface: I, theoretical considerations for single fibre pull-out tests' Composites Sci and Techno122 (1985) pp 33-42 31 Chua, P.S. and Piggott, M.R. 'The glass fibre-polymer interface: II, work of fracture and shear stresses' Composites Sci and Techno122 (1985) pp 107-119 32 Chua, P.S. and Piggott, M.R. 'The glass fibre-polymer interface: III, pressure and coefficient of friction' Composites Sci and Techno122 (1985) pp 185-196 33 Chua, P.S. and Piggott, M.R. 'The glass fibre-polymer interface: IV, controlled shrinkage polymers' Composites Sci and Technol 22 (1985) pp 245-258 34 Piggott, M.R. 'Debonding and friction at fibre-polymer interfaces. I, criteria for failure and sliding' Composites Sci and Technol30 (1987) pp 295-306 35 Penn, L.S. and Lee, S.M. 'Interpretation of experimental results in the single pull-out filament test' J Composites Technol Res 11 No 1 (Spring 1989) pp 23--30 36 Morrison, J.K., Shah, S.P. and Jenq, Y.S. 'analysis of fiber debonding and pull-out in composites' J Engng Mech 114 No 2 (Feb 1986) pp 277-29 37 Sigl, L.S. and Evans, A.G. 'Effects of residual stress and frictional sliding on cracking and pull-out in brittle matrix composites' Mech Mater8 (1989) pp 1-12 38 Sling, H. and Shah, S.P. 'Failure of fibre reinforced composites by pull-out fracture' J Mater Sci 21 (1986) pp 953-957 39 Atkinson, C., Avila, J., Betz, E. and Smelser, R.E. 'The rod pull out problem, theory and experiment' J Mech Phys Solids 30 No 3 (1982) pp 97-120 40 Bowling, J. and Groves, G.W. 'The debonding and pull-out of ductile wires from a brittle matrix' J Mater Sci 14 (1979) pp 431442 41 Kelly, A. 'Interface effects and the work of fracture of a fibrous composite' Proc Roy Soc London A 319 (1970) pp 95-111 42 Kawali, K. and Takeda, N. 'Analysis of shear tests of tapered lap adhesive joints and fiber pull-out tests based upon energy balance concept' Trans JSCM 4 No 1 (1978) pp 23-30 43 Ostrowski, J., Will, G.T. and Piggott, M.R. 'Poisson's stresses in fibre composites. I: Analysis' J Strain Anal 19 No 1 (1984) pp 43--49 44 Piggott, M.R., Sanadi, A., Chua, P.S. and Andison, D. 'Mechanical interactions in the interfacial region of fibre reinforced thermosets' in Composite Interfaces, Proc First Int Conf on the Composite Interface, May 1986 pp 109-121 45 Herrera-Franeo, P.J., Chiang, M. and Drzai, L.T. 'Bond strength measurement in composites--analysis of experimental techniques' submitted for publication to Composites Engng 46 Schuster, D.M. and Scala, B. 'The mechanical interactions of sapphire whiskers with a birefringent matrix' Trans Am Inst Min Engrs 230 (1964) pp 1635-1640 47 Wu, H.F. and Claypool, C.M. 'An analytical approach of the microbond test method used in characterizing the fibre-matrix interface' J Mater Sci Left 10 (1991) pp 269-272 48 Ting, T.C.T. and Hoang, P.H. 'Singularities at the tip of a crack normal to the interface of an anisotropic layered composite' Int J Solids Structures 20 (1984) pp 439-454 49 Carroi, B.J. 'Equilibrium conformations of liquid drops on thin cylinders under forces of capillarity. A theory for the roll-up process' Langmuir 2 (1986) pp 248-250 50 Ozzello, A., Grummon, D.S., Drzal, L.T., Kalantar, J., Loh, I.-H. and Moody, R.A. 'Interfacial shear strength of ion beam modified UHMW-PE fibers in epoxy matrix composites' Proc Mater Res Soc Syrup 153 (1989) pp 217-222 51 Gaur, U., Besio, G. and Miller, B. 'Measuring fiber/matrix adhesion in thermoplastic composites' Plast Engng (Oct 1989) pp 43-45 52 Penn, L.S. and Lee, S.M. 'Interpretation of the force trace for Kevlar/epoxy single filament pull-out tests' Fibre Sci and Techno117 (1985) pp 91-97 53 Gaur, U. and Miller, B. 'Microbond method for determination of shear strength of a fiber/resin interface: evaluation of experimental parameters' Composites Sci and Techno134 (1989) pp 35-51 54 McAlea, K.P. and Besio, G.J. 'Adhesion between polybutylene terephthalate and E-glass measured with a microdebond technique' Polym Composites 9 No 4 (Aug 1988) pp 285-290
55 Penn, L.S., Tesoro, G.C. and Zhou, H.X. 'Some effects of surface-controlled reactions of Kevlar 29 on the interface in epoxy composites' Po(~m Composites 9 No 3 (June 1988) pp 184-191 56 Latour Jr., R.A., Black, J. and Miller, B. 'Fatigue behavior characterization of the fibre-matrix interface' J Mater Sci 24 (1989) pp 3616-3620 57 McAlea, K.P. and Besio, G.J. 'Adhesion kinetics ofpolyhutylene to E-glass fibers' J Mater Sci Lett 7 (1988) pp 141-143 58 Cox, H.L. 'The elasticity and strength of paper and other fibrous materials' BritJAppl Phys 3 No 1 (March 1952) pp 72-79 59 Rosen, B.W. 'Mechanics of composite strengthening' in Fiber Composite Materials (American Society for Metals, 1964) pp 37-75 60 Amirhayat, J. and Hearle, J.W.S. 'Properties of unit composites as determined by the properties of the interface. Part I: mechanism of matrix-fibre load transfer' Fibre Sci and Technol 2 No 2 (1969) pp 123-141 61 Theocaris, P.S. The Mesophase Concept in Composites (Springer-Verlag, 1987) 62 Asloum, El M., Nardin, M. and Schultz, J. 'Stress transfer in single-fiber composites: effect of adhesion, elastic modulus of fibre and matrix and polymer chain mobility' J Mater Sci 24 (1989) pp 1835-1844 63 Folkes, M.J. and Wnng, W.K. 'Determination of interfacial shear strength in fibre-reinforced thermoplastic composites' Polymer 28 (July 1987) pp 1309--1314 64 Lhotellier, F.C. and Brinson, H.F. 'Matrix-fiber stress transfer in composite materials: elasto-plastic model with an interphase layer' Composite Struct 10 (1988) pp 281-301 65 Ran, V. and Drzal, L.T. 'The dependence of interfacial shear strength on matrix and interphase properties' Polym Composites 12 No 1 (Feb 1991) pp 48-56 66 Drzal, L.T. 'Intrinsic material limitations in using interphase modifications to alter fiber-matrix adhesion in composites' Mater Res Soc Syrup Proc (Materials Research Society, 1990) 170 67 Termonia, Y. 'Theoretical study of the stress transfer in single fibre composites' J Mater Sci 22 (1987) pp 504-508 68 Galiotis, C., Melanitis, N., Tetlow, P.L. and Davies, C.K.L. 'Interfacial shear stress mapping in model composites' in Proc 5th Tech Conf(American Society for Composites, 12-14 June 1990) 69 Verpoest, 1., Desaeger, M. and Keunings, R. 'Critical review of direct micromechanical test methods for interfacial strength measurements in composites' in Controlled lnterphases in Composite Materials, Proc Third Int Conf on Composite Interfaces edited by H. Ishida (Elsevier, NY, 1990) pp 653-666 70 Whitney, J.M. and Drzal, L.T. "Axisymmetric stress distribution around an isolated fiber fragment' in Toughened Composites, ASTM STP 937 (American Society for Testing and Materials, 1987) pp 179-196 71 Soh, A.K. 'On the determination of interfacial stresses in a composite material' Strain 21 No 4 (Nov 1985) pp 163-172 72 Carrara, A.S. and McGarry, F.J. 'Matrix and interface stresses in a discontinuous fiber composite model' J Composite Mater 2 No 2 (April 1968) pp 222-243 73 Pu, S.L. and Sadowski, M.A. 'Strain gradient effects on force transfer between embedded microflakes' J Composite Mater 2 No 4 (April 1968) pp 138-151 74 MacLanghlin, T.F. and Salkind, M.J. 'Effect of fibre geometry on stress in fibre-reinforced composite materials' Watervliet Arsenal Rep WVT-6521 (A D-616520) (1965) 75 Iremonger, M.J. and Wood, W.G. "Stresses in a composite material with a single broken fibre' J Strain Analysis 2 No 3 (1967) pp 239-245 76 Ko, W.L. 'Finite element microscopic stress analysis of cracked composite systems' J Composite Mater 12 (1978) pp 97-115 77 Netravali, A.N., Henstenburg, R.B., Phoenix, S.L. and Schwartz, P. 'Interfacial shear strength studies using the singlefilament composite test. I: experiments on graphite fibers in epoxy' Polym Composites 10 No 4 (Aug 1989) pp 226-241 78 Henstenburg, R.B. and Phoenix, S.L. 'Interfacial shear strength studies using the single-filament composite test. Part II: a probability model and Monte Carlo simulation' Polym Composites 10 No 5 (Dec 1989) pp 7-13 79 Bascom, W.D. and Jensen, R.M. 'Stress transfer in single fiber resin tensile tests' J Adhesion 19 (1986) pp 219-239 80 Wimolkiatisak, A.S. and Bell, J.P. 'Interfacial shear strength and
26
81 82
83
84 85 86 87 88 89
90
91
failure modes of interphase-modified graphite-epoxy composites' Polym Composites 10 No 3 (June 1989) pp 162-172 Fraser, W.A., Ancker, F.H., Dibenedetto, A.T. and Elbirli, B. 'Evaluation of surface treatments for fibers in composite materials' Polym Composites 4 No 4 (Oct 1983) pp 238-248 Asloum, El M., Donnet, J.B., Guilpain, G., Nardin, M. and Schultz, J. 'On the determination of the tensile strength of carbon fibers at short lengths' J Mater Sci 24 (1989) pp 35043510 Netravali, A.N., Topoleski, L.T.T., Sachse, W.H. and Phoenix, S.L. 'An acoustic emission technique for measuring fiber fragment length distributions in the single-fiber-composite test' Composites Sci and Techno135 (1989) pp 13-29 Waterbury, M.C. and Drzal, L.T. 'On the determination of fiber strengths by in situ fiber strength testing' J Composites Technol Res 13 No 1 (Spring 1991) pp 22-28 Ashhee, K.H.G. and Ashbee, E. 'Photoelastic study of epoxy resin/graphite fiber load transfer' J Composite Mater 22 (July 1988) pp 602--615 Waterbury, M.C. PhD Dissertation (Michigan State University, 1991) Mandell, J.F., Chen, J.H. and McGarry, F.J. in Proc35th Conf of RP/Composites Institute (Society of the Plastics Industry, 1980) paper 26D Grande, D.H. 'Microdebonding test for measuring shear strength of the fiber/matrix interface in composite materials' MS Thesis (Massachusetts Institute of Technology, June 1983) Mandeil, J.F., Grande, D.H., Tsiang, T.H. and McGarry, F.J. 'A modified microdebonding test for direct in-situ fiber/matrix bond strength determination in fiber composites' Res Rep R84-3 (Massachusetts Institute of Technology, Dec 1984) Grande, D.H., Mandell, J.F. and Hong, K.C.C. 'Fibre-matrix bond strength studies of glass, ceramic and metal matrix composites' J Mater Sci 23 (1989) pp 311-328 Netravali, A.N., Stone, D., Ruoff, S. and Topoleski, T.T.T. 'Continuous microindentor push through technique for measur-
92 93
94 95
96 97 98
99
ing interfacial shear strength of fiber composites' Composites Sci and Techno134 (1989) pp 289-303 Tse, M.K. US Patent 4 662 228 (1987) Caldwell, D.L., Babbington, D.A. and Johnson, C.F. 'Interfacial bond strength determination in manufactured composites' in Interracial Phenomena in Composite Materials edited by F.R. Jones (Butterworth, 1989) pp 44-52 Coidwell, D.L. and Cortez, F.M. Modern Plastics (Sept 1988) p 132 Petit, PAt. 'A simplified method of determining the inplane shear stress-strain response of unidirectional composites' ASTM STP 460 (American Society for Testing and Materials, 1969) p 63 Rosen, B.W. 'A simple procedure for experimental determination of the longitudinal shear modulus of unidirectional composite materials' J Composite Mater 6 (1972) pp 552-554 Whitney, J.M., Daniel, 1.M. and Pipes, R.B. 'Experimental mechanics of fiber reinforced composite materials' Monograph 4 (Society for Experimental Stress Analysis, 1982) Walrath, D.E. and Adams, D.F. 'Analysis of the stress state in an Iosipescu shear test specimen' Department Rep UWME-DR301-102-1 (Composite Materials Research Group, Department of Mechanical Engineering, University of Wyoming, Laramie, June 1983) Rao, V., Herrera-Franco, P., Ozzello, A.D. and Drzal, L.T. 'A direct comparison of the fragmentation test and the microbond pull-out test for determining the interfacial shear strength' J Adhesion 34 (1991) pp 65-77
A U THORS
The authors are with the Composite Materials and Structures Center, Michigan State University, East Lansing, MI 48824-1326, USA.
COMPOSITES
. JANUARY
1992
27

Comparison of Methods For The Measurement of Fibre - Matrix Adhesion in Composites

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Comparison of Methods For The Measurement of Fibre - Matrix Adhesion in Composites

Uploaded by

Copyright:

Available Formats

Comparison of methods for the measurement of fibre/matrix adhesion in composites

Polymer of different properties

Fibre morphology Bulk fibre

Fibre/matrix interphase Fig. 1 Characteristics of the fibre/matrix interphase in a composite material

COMPOSITES. JANUARY 1992

\ Microdroplet Fibre diameter (d)

MICROBOND TECHNIQUE Background

COMPOSITES. JANUARY 1992

Theoretical considerations 1. Pull-out technique

which, using Equation (1), results in:

Effect of pressure on m a x i m u m pull-out

Typical force/displacement plot from a pull-out experi-

1;i--[ 1-{'- 2*~xp(I+2L) ]Efr

COMPOSITES. JANUARY 1992

COMPOSITES. JANUARY 1992

Experimental apparatus and procedure

~iiiiii ~i'iii~ii:!:ili ..........

COMPOSITES. JANUARY 1992

SINGLE-FIBRE FRAGMENTATION TECHNIQUE Background

+~ Thin strip of polymer film

Procedure to form thermoplastic resin droplets on the

Fig. 12 Schematic representation of the single-fibre fragmentation process

COMPOSITES . JANUARY 1992

where d is the fibre diameter for a circular fibre cross-

Fig. 13 Micromechanical model ofthe single-fibre fragmentation test TM

where (O<.r<.R). The interfacial shear stress is given by the expression:

Photoelastic and finite element stress analysis

+ [(lWVm)Em--E2f-Vl2t'Elf]} Co 2gfGm C2= (gf+Gm) (Vl2f"Vm)-~ [(i

COMPOSITES . JANUARY 1992

Fig. 14 Distribution of axial stress along fibre fragment

Statistical characterization of the critical fibre fragment length distribution

Measurement of the critical fibre fragment length

Experimental apparatus and procedure

MICRODEBONDIMICRO-INDENTA TION TECHNIQUE

0.8 1.0 I / ~ /rf I I

Compression Normalized stresses

Fig. 29 Interfacial stress d i s t r i b u t i o n f o r a c a r b o n f i b r e / e p o x y m a t r i x s y s t e m 9

1.0 I 0.15 0.10 t

I I 0.20 0.25 0.30 ~,

Fig. 28 Interfacial stress d i s t r i b u t i o n f o r H M U c a r b o n f i b r e s in b o r o s i l i c a t e glass m a t r i x o b t a i n e d u s i n g the m i c r o - i n d e n t a t i o n t e c h n i q u e 9

Matrix~]] Fibre ~ 0.30

|j',/ FEM model

! ! t ' ~z Hertzian, homogeneous

Experimental apparatus and procedure

2. Iosipescu shear test specimen

Short-beam shear test

Composite laminate techniques for shear characterization

1. (+45)s tensile test

losipescu shear test

Short beam interlarninar shear test

Schematic representation of laminate c o m p o s i t e shear

Summary of interfacial shear strength values

COMPOSITES. JANUARY 1992

ADVANTAGES AND LIMITATIONS Microbond technique

You might also like