Fractographic study of transverse cracks in a fibre composite

Transverse fracture of unidirectional fibre composites was studied in a model glass/epoxy composite in which 1 mm-diameter rods had been used in place of fibres. The fracture surface resulting from transverse cracking in this model system was studied by scanning electron microscopy (SEM). The interaction of the crack with the epoxy matrix resin and the glass rods was the following: Cracks in the resin appeared to have effected a debonding at the glassmatrix interface before reaching the glass. The debonding then propagated along the interface and induced secondary cracks ahead of the primary debonding crack. The confluence of the secondary and primary cracks resulted in sharp ridges being formed on the matrix resin surface, produced by plastic deformation of the rigid epoxy resin. These appeared as a field of parabolic marks. Considering the brittleness of the resin, the amount of plastic deformation indicated by the ridges was astonishing. As the debonding continued around the glass rod, a transverse corrugated texture developed on the resin surface, again produced by plastic deformation. Finally, the cracks reentered the matrix from small patches of polymer adhering especially strongly to the glass surface. The overall fracture energy of transverse cracking of unidirectional fibre composites is suggested to consist, therefore, of the following elements in addition to crack propagation in the matrix resin: (a) the glass-resin debonding before the incoming cracks reach the glass, (b) the initiation of secondary cracks or debonds at the interface, (c) the plastic deformation in generating the ridges on the rigid resin surface, appearing both as the paraboloids and the transverse corrugation, and (d) cracking of the matrix reinitiated at the opposite side of the glass. The use of an enlarged glass reinforcement in this study provided a more direct observation of the properties of transverse crack propagation in composite materials than would have been possible with the small, roughly 10m fibres.     

Hybrid C/Pet/epoxy composites

Hybrid carbon/PET/epoxy composites were prepared by a wet layup technique and characterized by standard methods of fracture and impact tests. The influence of PET fibre surface treatment with polyfunctional amine on the composite properties has been studied in order to determine the role of the fibre/matrix interface. The state of the fibre/matrix interface has been examined by several techniques, such as SEM, differential scanning calorimetry (DSC) and contact angle measurements. A significant hybrid effect of the PET/aminated fibres on the curing reaction and T _g of the epoxy matrix in the composites was established. It is probably due to the reaction between the amine groups of the fibre surface and the epoxy component of the resin, as determined by DSC.     

The energy of crack propagation in carbon fibre-reinforced resin systems

The energy expended during controlled crack propagation in unidirectionally reinforced composites of carbon fibre in a brittle resin matrix has been evaluated in terms of the energy dissipated during fibre-snapping, matrix-cracking and fibre pull-out. The work of fracture, _F, is found to depend principally on the frictional shear stress at the fibre/resin interface opposing pulling out of broken fibres. Differences in _F for carbon fibre/resin composites exhibiting a range of interfacial shear strengths and void contents have been explained with reference to variations in fracture surface topography of the fibrous composites. The effect of environment on properties of the interface and work of fracture was also investigated. The energy required to propagate a crack has been compared with the energy for fracture initiation, _I, using a linear elastic fracture mechanics approach. It was found that fibre pull-out energy is the principal contribution to _F, and _I is similar to the elastic strain energy release rate at the initiation of fracture of a brittle, orthotropic solid. For crack propagation parallel to fibres, _F and _I are similar and not unlike the fracture surface energy of the resin alone. The strength of the interface is important only in so far as it affects the value of _I.     

What happens after a fibre breaks — pull-out or resin cracking?

We consider tensile fracture of a specimen consisting of a single rigid fibre embedded in a cylindrical block of a linearly-elastic resin. When the fibre breaks, two possible modes of failure can occur. A circular crack may propagate outwards into the resin, leading to fracture of the specimen. Alternatively, a cylindrical crack can propagate along the fibre-matrix interface, starting from the break in the fibre, leading to fibre pull-out. The question is: which mode of failure will occur in practice? Finite-element analysis is used here to calculate the pull-out force and the force causing growth of a circular crack outwards into the resin, for samples containing fibres of different radius. A general criterion is obtained to predict the mode of failure. Even for samples with perfect adhesion between resin and fibre, pull-out of the fibre is expected when the fibre radius is less than about one-fifth of the sample radius. For fibres of larger radius, either pull-out or resin cracking can take place, depending on the relative levels of interfacial fracture energyG _a and resin fracture energyG _c.     

Tensile properties of real unidirectional Kevlar/epoxy composites

Mechanical behaviour, tensile strength and failure modes in real unidirectional Kevlar/epoxy composites, loaded parallel to the fibres, at volume fraction (V_f) range 0.26–0.73, were investigated. It was found that the measured tensile strengths deviated from the expected values calculated from the Rule of Mixture. The deviation, which was minimal at V_f of about 0.5, was mainly due to geometrical deficiencies typical of real composites. At V_f<0.5 it could be explained by non-homogeneous fibre spread and distribution of fibres. At V_f>0.5 the deviation was explained by the increasing lack of matrix between some adjacent fibres and by squeezing of fibres. The initial part of loading was typified by straightening out of non-axial fibres, accompanied by fibre/matrix debonding. The straightening process was completed at a stress level of about 0.6–0.7 of the composite strength. Matrix damage began at this stress level and continued to develop up to final failure. Failure of Kevlar fibres was noted to occur only at an extremely short loading interval coinciding with the catastrophic final failure. This was due to the small scatter of Kevlar fibre strength.     

Compressive fracture behavior of 3D needle-punched carbon/carbon composites

Compressive fracture behavior under transverse and longitudinal compressive loading are determined for 3D needle-punched carbon/carbon (C/C) composites with single rough laminar (RL) pyrocarbon matrix or dual matrix of RL pyrocarbon and resin carbon. The results of Weibull statistics analysis indicate that scale parameter σ₀ of transverse and longitudinal compression of the composites with single matrix are 153.41 and 94.26 MPa, and σ₀ of the composites with dual matrix are 205.16 and 105.33 MPa, respectively. The mean compressive strength of both composites is nearly equal to σ₀ under each experimental condition. Failure modes of both composites under transverse and longitudinal compressive loading are shear and extension, respectively. Both composites exhibit quasi-ductile fracture behavior under transverse compression. Many small fragments of fibers and matrix carbon on the fracture surface of the composites are observed for single matrix composites. And the fiber bundle breakage with extensive debonding occurs for dual matrix composites. Under longitudinal loading, the composites with single matrix show quasi-ductile fracture behavior and delamination and splitting of non-woven long carbon fiber cloth layers are observed. The composites with dual matrix exhibit catastrophic failure behavior and crack runs through the composites along compressive loading direction.     

Interfacial failure in poly(p-phenylene benzobisoxazole) (PBO)/epoxy single fibre pull-out specimens

The pull-out behaviour of poly(p-phenylene benzobisoxazole) fibres from an epoxy resin has been shown to follow that predicted by the elastic stress transfer shear-lag model at low applied strains, but at higher matrix strains a partial debonding model was more suitable. Debonding of the fibre/matrix interface led to interfacial failure where only friction resisted fibre extraction. Raman spectroscopy was able to quantify this level of friction and together with in situ optical microscopy proved an excellent method for the close monitoring of the frictional pull-out process. The effect of fibre surface treatment was also studied. The interfacial shear stress values from the heat-treated and corona-treated fibres showed only small differences. The failure processes were examined further using scanning electron microscopy and clean fibre pull-out was observed with the heat-treated fibre whereas fracture of the free fibre occurred with the corona-treated fibre.     

Experiments on shear deformation,debonding and local load transfer in a model graphite/glass/epoxy microcomposite

Recent statistical theories for the failure of polymer matrix composites depend heavily on details of the stress redistribution around fibre breaks. The magnitudes and length scales of fibre overloads as well as the extent of fibre/matrix debonding are key components in the development of longitudinal versus transverse crack propagation. While several theoretical studies have been conducted to investigate the roles of these mechanisms, little has been substantiated experimentally about the matrix constitutive behaviour and mechanisms of debonding at the length scale of a fibre break. In order to predict the growth of transverse and longitudinal cracks using the same micromechanical model, we microscopically observed the epoxy shear behaviour around a single fibre break in a three-fibre microcomposite tape. The planar specimens consisted of a single graphite fibre placed between two larger glass fibres in an epoxy matrix. The interfibre spacing was less than one fibre diameter (<6 m) in order to reflect the spacing between fibres found in typical composites. The epoxy constitutive behaviour was modelled using shear-lag theory where the epoxy had elastic, plastic, and debond zones. The criteria for debonding were modified from conventional shear-lag approaches to reflect the orientational hardening in the epoxy network structure. The epoxy, which is brittle in bulk, locally underwent a shear strain of about 60% prior to debonding from the fibre.     

Investigation of the mechanical behaviour of magnesium composites

Stress/strain and fracture toughness behaviour of a commercial heat-treatable magnesium alloy reinforced with up to 20 volume% short alumina fibres was studied at room and elevated temperatures. Microscopic examination of the composites, which were prepared by conventional squeeze casting, revealed damage of a small portion of the fibres during the infiltration process. Sufficient chemical reaction between the matrix alloy and alumina reinforcement tends to produce a good bond at the fibre/matrix interface. The tensile-related properties of the composites increased at room and elevated temperatures with increasing content of the reinforcement. The ductility and fracture toughness of the composites decreased at room temperature with increasing reinforcement content. While failure strains of the composites were slightly improved at higher testing temperatures, the fracture toughness decreased significantly as the testing temperature exceeded 100°C. Examination of the fracture surfaces of specimens tested at room temperature showed a mixed mode fracture appearance with predominantly brittle cleavage fracture. The fracture surfaces of specimens tested at temperatures above 100°C revealed increasing fibre/matrix interface debonding and fibre pull-out with increasing testing temperature. Micromechanism examinations of crack initiation and propagation indicated that the fracture process of the composites may be matrix controlled.     

Photoelastic study of the durability of interfacial bonding of carbon fibre-epoxy resin composites

The physical techniques of polarizing microscopy, including the quantitative measurements of small optical retardations, have been used to investigate elastic fields adjacent to short carbon fibres in epoxy resin composites. The elastic fields associated with shear stress distribution along the fibre-matrix interface have been employed to monitor the initiation of interface debonding during hot (100 °C) water uptake. By examining the development of stress birefringence during resin swelling in the resin adjacent to individual fibres, the differences in the durability of interfacial bonding and the fibre failure modes for differently coated fibres have been obtained. The results show that the state of self-stress in model composites, comprising a single carbon fibre in a film of epoxy resin, can, by immersion in hot distilled water, be enhanced to such an extent that the axial tension in the fibre can be sufficient to initiate fibre fracture. The results also show that, for fibres that have been given certain proprietary surface treatments, the fibre fractures by different failure modes.     

Fracture toughness and reliability in high-temperature structural ceramics and composites: Prospects and challenges for the 21st century

The importance of high fracture toughness and reliability in Si₃N₄, and SiC-based structural ceramics and ceramic matrix composites is reviewed. The potential of these ceramics and ceramic matrix composites for high temperature applications in defence and aerospace applications such as gas turbine engines, radomes, and other energy conversion hardware have been well recognized. Numerous investigations were pursued to improve fracture toughness and reliability by incorporating various reinforcements such as particulate-, whisker-, and continuous fibre into Si₃N₄ and SiC matrices. All toughening mechanisms, e.g. crack deflection, crack branching, crack bridging, etc essentially redistribute stresses at the crack tip and increase the energy needed to propagate a crack through the composite material, thereby resulting in improved fracture toughness and reliability. Because of flaw insensitivity, continuous fibre reinforced ceramic composite (CFCC) was found to have the highest potential for higher operating temperature and longer service conditions. However, the ceramic fibres should display sufficient high temperature strength and creep resistance at service temperatures above 1000°C. The greatest challenge to date is the development of high quality ceramic fibres with associate coatings able to maintain their high strength in oxidizing environment at high temperature. In the area of processing, critical issues are preparation of optimum matrix precursors, precursor infiltration into fibre array, and matrix densification at a temperature, where grain crystallization and fibre degradation do not occur. A broad scope of effort is required for improved processing and properties with a better understanding of all candidate composite systems.     

Role of fibre surface—matrix combination in carbon fibre reinforced epoxy composites

The effect of surface treatment of carbon fibres with concentrated as well as dilute nitric acid on the mechanical properties of carbon fibres has been reported. The role of the fibre—matrix interface in carbon fibre reinforced epoxy resin composites has been studied. Composites have been made both with untreated and surface treated carbon fibres and epoxy resin Araldite LY556 with different hardeners. Mechanical properties as well as fracture behaviour of these composites suggest that it is the physical interlocking between the fibres and the matrix, along with some chemical bonding between the two, and not the pure chemical bonding which yield better composites.     

The impact strength of fibre composites

Two models have been developed which predict the crack initiation energy, notched impact strength and unnotched impact strength of fibre composites. One is applicable to composites containing short fibres and the other to composites containing long fibres. Data obtained with randomly oriented short fibre composites were consistent with the one model. The other model has been verified using composites containing uniaxially oriented long fibres and long fibres oriented randomly in a plane. The success of the model demonstrates that the high notched impact strength with long fibres is due to the redistribution of stress away from the stress concentrating notch, the extra stress that can be held by the fibre relative to the matrix and the work required to pull fibres out of the matrix during crack propagation. The parameters which have been shown to control the fracture energy are composite modulus, fibre length, fibre volume fraction, effective fibre diameter, fibre tensile strength and the coefficient of friction during fibre pull-out from the matrix. The matrix toughness on the other hand usually has no effect at all for composites containing fibres randomly oriented in two dimensions and only a minor effect in exceptional cases. The shear strength of the fibre-matrix bond has only an indirect effect in that it controls the number of fibres which pull out rather than fracture.     

Fatigue damage development in Al2O3/Al composite

Fatigue damage development in two aluminium matrix (Al7SiO.6Mg and Al5Si-3Cu1Mg) composites reinforced with discontinuous Al₂O₃ fibres has been monitored by means of acoustic emission (AE). The AE signals (RMS) recorded during the tests clearly exhibit three distinct stages which correspond to crack initiation, dominant crack formation and stable propagation. Generally speaking, the cracks initiated at a high load level form close together and a dominant crack forms easily. By contrast, at a low load, initiated cracks are widely separated and the formation of a dominant crack is difficult. If there are large defects in the composite, the first stage is absent, even at low load. In the first stage, little change in microstructure and modulus of the composite is observed; in the second, fibre fracture, interface debonding and matrix cracking occur and there are often sinusoidal cracks in the matrix; in the last stage, the principal characteristic is stable propagation of the dominant crack. The degradation of the elastic modulus of the composite in the last two stages is small.     

Influence of interfacial stress transfer on fatigue crack growth in SiC-fibre reinforced titanium alloys

An important damage mechanism during fatigue of unidirectional SiC-fibre reinforced titanium alloys is the formation of matrix cracks transverse to the fibre direction. Due to the relatively low fibre/matrix bond strength these matrix cracks initially do not break the fibres, so that matrix cracks bridged by fibres develop. It is shown experimentally, that the strong drop in fatigue strength is caused by the formation of a bridged crack of a critical size and the crack propagation rate (da/dN) for a single load level has been determined. A prediction of da/dN on the basis of finite element calculation of the stress intensity factor range of the bridged matrix crack ΔK_m and the ΔK_m–da/dN relationship of the used titanium alloy (Timetal 834) has been performed. Calculation of ΔK_m assuming a negligible fibre/matrix bond strength and considering shear load transfer at the fibre/matrix interface due to Coulomb friction (coefficient of friction μ=0.5 and μ=0.9) led to a large discrepancy between the measured and predicted crack growth rate. It can be concluded, that the assumed conditions of stress transfer at the fibre/matrix interface neglecting bonding is the reason for this discrepancy.     

Some aspects of crack growth and failure in fibre reinforced composites

A theoretical analysis, previously developed to deal with the machanics of matrix cracking in unidirectional composites and with transverse ply cracking in cross ply laminates, has been developed further to deal with the tensile failure of unidirectional fibrous composites in with the fibres have a known distribution of strengths. It is proposed that, under the application of a tensile load, stable transverse cracks are formed which originate from regions of initial damage and which become unstable at some critical strain value. The model takes account of various parameters including the interfacial fibre/matrix debonding energy, the residual frictional shear strength of the debonded interface and the elastic properties of fibres and matrix. Comparisons are made between the predictions of the model and the observed failing strains of the 0° plies in carbon fibre polymer matrix laminates. The relevance of the model to the study of delayed fracture in fibrous composites is discussed. The modification of this model, previously developed to describe crack growth in the transverse plies of 0°/90° laminates, is used to predict the initial cracking strains for a wide range of CFRP laminate geometries and initial crack sizes. Some aspects of the mechanics of crack extension across interply interfaces are discussed.     

An evaluation of acoustic emission from fibre-reinforced composites

An analysis of acoustic emission (AE) from model composites consisting of a single aramid fibre and different epoxy matrix systems has been carried out to identify the source of acoustic emission. The AE activity was observed in a narrow range of strain when fibre fracture occurred, whereas in a relatively wide range of strain, debonding occurred at the fibre-matrix interface. Ion-etched fibres showed a good adhesion of the fibres to the matrix so as to produce fibre fracture in place of interfacial debonding. The total number of AE events has one to one correspondence with the number of broken fibres. The effect of surface treatment and matrix systems on the shear fracture strength between the fibre and matrix were described based on the critical length of the broken fibres using AE results.     

Three techniques of interfacial bond strength estimation from direct observation of crack initiation and propagation in polymer–fibre systems

Three techniques of bond strength determination in micromechanical tests—fibre strain profile analysis by means of Raman spectroscopy, “kink” force determination in a traditional pull-out test, and crack length monitoring in a microbond test—were used for investigation of interfacial debonding in epoxy–glass fibre and epoxy–aramid fibre systems. Crack propagation was characterised by local interfacial parameters—critical energy release rate, G_ic, and ultimate interfacial shear strength (IFSS), τ_ult. The comparison of the results showed good agreement both between different techniques and between stress-based and energy-based failure criteria. Sizing of glass fibres caused more pronounced variations in the IFSS than for aramid fibres due to different interfacial failure patterns. The strength of “real” epoxy–glass composites with sized and unsized fibres correlates well with the bond strength determined from the micromechanical tests.     

The fracture toughness of some wood-ice composites

Tests have been performed on a number of ice composites, using wood based products as reinforcing agents, to determine the fracture toughness of such composites. The results show that for small amounts of reinforcing material (between 5 and 20%) increases in fracture toughness, over the value for freshwater ice, of between 500 and 2000% can be obtained, giving values of K_Ic of between 0.69 and 1.92 MPa√ m.Of the reinforcing materials used, the largest toughness at a given weight percentage reinforcement was obtained from those samples in which the reinforcing fibres had the smallest diameter. This trend is consistent with a simple model of energy dissipation at cracks in composites, and suggests that the energy absorbing mechanisms which produce the toughening in this instance might be fibre pull-out resistance and fibre debonding. However, further work is needed to determine what other toughening mechanisms may also be of significance.     

Penetration impact resistance of hybrid composites based on commingled yarn fabrics

The penetration impact resistance of hybrid composites based on commingled yarn fabrics was investigated. The commingled yarn fabrics were composed of E-glass fibres (GF) and thermoplastic fibres blended together within the fibre bundles. Various thermoplastic fibres such as polypropylene (PP), polyamide (PA) and modified polyethylene terephthalate (mPET) were studied. Various resin matrices with different cure cycles were studied such as Quickcure polyester, Cycom X823 RTM epoxy, and Shell Epikote 828 epoxy resin. Depending on the crystalline melting temperature (T_m) of the thermoplastic fibres, the hybrid composites can be categorised as fibre-hybrid composites or matrix-hybrid composites. Fibre-hybrid composites refer to those in which the thermoplastic fibres remain in the fibre form after curing, for example the GF–PP and GF–PA hybrid composites. For matrix-hybrid composites, the thermoplastic fibres melt and dissolve into the thermosetting matrix during curing such as the GF-mPET hybrid composites. The results from the penetration impact showed that the total absorbed energy of the fibre-hybrid composites were significantly higher than for the plain glass composites. Plastic deformation in the thermoplastic fibres is the key factor that improves the absorbed energy of the hybrid composites. When the thermoplastic fibres dissolved into the thermosetting matrix as in matrix-hybrid composites, the total absorbed energy was similar to that of the plain glass fibre composites. This suggests that the total absorbed energy is dependent on the properties of the fibres rather than the matrix. However, the fibre-hybrid composites showed slight differences in the total absorbed energy for different matrices. The differences are thought to be related to the differences in the bonding between the thermoplastic fibres and the thermosetting matrix which have yet to be investigated.