Compressive failure of fibre-reinforced composites: buckling,kinking, and the role of the interphase

Recent experimental studies of compressive failure in fibre-reinforced polymeric composites have been analysed. It is shown that the parametric basis for most compressive strength models, i.e. pure plastic buckling controlled by matrix shear strength and initial fibre misorientation, is probably incomplete. It is argued that, instead, failure is triggered by the initiation of an unstable kink band prior to buckling instability, and that additional parameters (interfacial shear stress/strain; fibre strength) are responsible for this transition in mechanisms.     

A Graphical Method Predicting the Compressive Strength of Toughened Unidirectional Composite Laminates

The in-plane shear and compressive properties of unidirectional (UD) HTS40/977-2 carbon fibre-toughened resin (CF/TR) laminates are investigated. Scanning Electron microscopy (SEM) and optical microscopy are used to reveal the failure mechanisms developed during compression. It is found that damage initiates by fibre microbuckling (a fibre instability failure mode) which then is followed by yielding of the matrix to form a fibre kink band zone that leads to final fracture. Analytical models are briefly reviewed and a graphical method, based on the shear response of the composite system, is described in order to estimate the UD compressive strength. Predictions for the HTS40/977-2 system are compared to experimental measurements and to data of five other unidirectional carbon fibre reinforced polymer (CFRP) composites that are currently used in aerospace and other structural applications. It is shown that the estimated values are in a good agreement with the measured results.     

The mechanical properties and fracture behaviour of unidirectionally reinforced Nylon 6

The tensile and shear properties of Nylon 6 polymerizedin situ around unidirectionally aligned carbon and glass fibres have been investigated and the fracture behaviour characterized by optical and scanning electron microscopy. The tensile strengths are found to lie within the limits predicted by the law of mixtures and deviations from the predicted strengths have been correlated with fibre type and surface treatment. The shear strength values follow the same trend and an important mode of fracture in bending is shown to be the compressive failure which accompanies a yield drop in the load deflection curve. Depending upon the fibre type and the properties of the matrix this compressive damage need not lead to catastrophic failure of the composite as, in certain cases, the matrix can undergo substantial deformation before failure.     

Theoretical and experimental evaluation of the Iosipescu shear test

Use of the Iosipescu shear test for measurement of shear properties of unidirectional laminae has been studied both analytically and experimentally. The intralaminar shear strength and shear stiffness of glass-reinforced polyester material have been measured using specimens with two different fibre orientations. Acoustic emission has been monitored and a fractographic study carried out. A finite element analysis has been conducted to evaluate the stress distribution within the specimen, assuming isotropic and orthotropic elastic properties of the material. There is a complicated stress distribution in the specimen, particularly in the vicinity of each notch root, depending on the elastic properties of the composite. The shear stress region in the specimen gauge section is almost uniform, though small normal compressive stresses exist. The experimental results have shown that the measured shear modulus does not depend on reinforcement orientation. However, it has been observed that different failure modes occur in each case. This results in a change in apparent shear strength of the composite with fibre orientation. Some explanations of these differences have been found in a detailed analysis of the local stresses at the roots of the notches. It is considered that the presence of tensile stresses in this area is primarily responsible for the apparent reduction in the shear strength.     

A fibre direction compressive failure criterion for long fibre laminates at ply scale,including stacking sequence and laminate thickness effects

Long fibre laminate compressive failure is due to a microbuckling instability which leads to a kink band and a brittle failure of the fibres. This failure mechanism is well known, but more or less pertinently explained in the literature. Some references also showed that local microbuckling instability depends on parameters that belong to the scale of the elementary ply, like thickness and corresponding lay-up. The compressive strength of the unidirectional ply is therefore no more an intrinsic material property, but results from a structural effect of the design. In this paper, the so-called “structure effect” is included in a simple way as an analytical formula in the phenomenological compressive failure criterion which was initially presented by Budiansky and Fleck works. The criterion presented is expressed analytically for unidirectional composite and stands for the local compressive failure strength at ply scale in fibres direction.     

Schubspannungsbomben in Faserverbunden

Shear Bombs in Fibre Composites Despite an optimum external shape non‐load adapted internal fibre orientation can lead to the formation of shear cracks where crossing tension‐compression principal stress trajectories create localized shear peaks. Trees are subject to those failure because they cannot re‐arrange their fibres after wood formation. Bones can adjust their micro‐structure to changing load conditions and in this way can better control shear failure. The engineer working with fibre composites should be alert to avoid fibre arrangements not following the force flow. Localized shear zones may also form near notches similar to normal notch stresses, however they are not always situated at the contour line of the notch.     

The effect of transverse shear on the longitudinal compressive strength of fibre composites

Experimental work on glass/epoxy composites shows that the compressive strength is sensitive to the method of gripping, that the failure mode in compression varies with fibre volume fraction, and that bending of the specimen may occur as a result of misalignment. Some aspects of these observations are examined. The critical Euler buckling load is significantly reduced if transverse shear occurs. The buckling load depends on specimen dimensions and a good deal of scatter results from this. The predicted compressive strength taking into account the effect of transverse shear and specimen geometry includes the experimental results within a wide scatter band. The present analysis based upon the macro-buckling of the specimen, reproduces some predictions of compressive strength based upon the micro-buckling of fibres.     

Fracture mechanisms and failure analysis of carbon fibre/toughened epoxy composites subjected to compressive loading

This study investigates the failure mechanisms of unidirectional (UD) HTS40/977-2 toughened resin composites subjected to longitudinal compressive loading. A possible sequence of failure initiation and propagation was proposed based on SEM and optical microscopy observations of failed specimens. The micrographs revealed that the misaligned fibres failed in two points upon reaching maximum micro-bending deformation and two planes of fracture were created to form a kink band. Therefore, fibre microbuckling and fibre kinking models were implemented to predict the compressive strength of UD HTS40/977-2 composite laminate. The analysis identified several parameters that were responsible for the microbuckling and kinking failure mechanisms. The effects of these parameters on the compressive strength of the UD HTS40/977-2 composite systems were discussed. The predicted compressive strength using a newly developed combined modes model showed a very good agreement to the measured value.     

Property Optimisation in Fibre Metal Laminates

Fibre Metal Laminates (FMLs) are hybrid materials, which consist of thin metal sheets bonded together with alternating unidirectional fibre layers. This material concept has resulted in superior fatigue characteristics with respect to the metallic counterpart. Several static characteristics (specifically tension, shear, bearing, blunt and sharp notch behaviour) are however negatively influenced due to the fibre addition. This paper investigates the influence of constituent properties on these characteristics to define possible improvements. The available analytical models are reviewed and if necessary the specific failure mechanisms are described. Using these models and available test data, trend lines are obtained, which indicate the effects of the principal parameters and quantify potential improvements.     

Crack nucleation from a notch in a ductile material under shear dominant loading

We examine the nucleation of a crack from a notch under a dominant shear loading in Al 6061-T6. The specimen is loaded in nominally pure shear over the gage section in an Arcan specimen configuration. The evolution of deformation is monitored using optical and scanning electron microscopy. Quantitative measurements of strain are made using the 2nd phase particles as Lagrangian markers which enable identification of the true (logarithmic) strains to levels in the range of two. Electron microscopy reveals further that the 2nd phase particles do not act as nucleation sites for damage in the regions of pure shear deformation. The initial notch is shown to “straighten out”, forming a new, sharper notch and triggering failure at the newly formed notch. Numerical simulations of the experiment, using the conventional Johnson–Cook model and a modified version based on grain level calibration of the failure strains, reveal that it is necessary to account for large local strain levels prior to the nucleation of a crack in order to capture the large deformations observed in the experiment.     

Mechanics of Elastomer--Shim Laminates

The mechanics of laminates of elastomer and shims of high modulus material are reviewed. Such structures are often built to provide engineering components with specified, and quite different, stiffnesses in different modes of deformation. The shims may either be rigid or flexible, flat or curved, but are usually close to inextensible, being made of a high modulus material such as steel. On the other hand, rubber has an exceptionally low shear modulus, about one thousandth of its bulk modulus, so that shear of the rubber layers and flexure of the high modulus layers (if thin) are the dominant mechanisms of deformation of the composite. In comparison, extension of the layers and changes to their separation are highly constrained.
Modes of failure are addressed as well as force-deformation behaviour. For compression normal to the laminations, the shear in the rubber results in in-plane tension in the shims, possibly leading to tensile failure. For tension normal to the laminations, the elastomer can cavitate, which would relieve the shear in it and hence the in-plane compressive stress applied to the shim. In flexure, shear in the rubber can apply in-plane compressive stress to the shims and cause buckling failure.     

Measurement of the fracture toughness associated with the longitudinal fibre compressive failure mode of laminated composites

The fracture toughness associated with the fibre compressive failure was obtained from testing notched unidirectional carbon/epoxy four-point-bend specimens. Microscopy of failed specimens revealed that onset of damage was characterised by the formation of a single line of fibre breaks at approximately 45° to the plane of the initial notch. A micromechanical finite element model was used to investigate this failure scenario and it was concluded that the most probable cause of the damage morphology was compression-induced shear failure of the composite. An intrinsic material property in this case was deemed to be the mode II critical strain energy release rate associated with the initiation of the 45° crack. For IM7/8552, this was measured to be G_IIc = 4.5 ± 0.8 kJ/m².     

Compressive behaviour of large undamaged and damaged thick laminated panels

Both intact and impact-damaged laminated panels under in-plane compressive loading are investigated with a purpose-built anti-buckling support. The readings of back-to-back strain gauges of selected locations are used to deduce panel behaviour in addition to post-mortem observation. The compression failure of intact panels is found to be close to the potted end. The failure characteristics of impact-damaged panels are dependent slightly on composite systems although they all failed in compression in the impact-damaged region with a kink shear band passing through the mid-section. E-glass/polyester panels with a greater shear angle do not seem to involve global buckling like S-glass/phenolic panels with small shear angles. The fact that the region covered by a kink shear band from impact surface to the distal surface is considerably less than the delamination area suggests that the initiation of overall failure is due to the collective result of flexural stiffness reduction compounded by the local impact damage and the associated change of fibre curvature. As a result, the residual compressive strengths are reduced significantly. Further outward propagation of the existing delamination(s) along the mid-section during loading is visible only for E-glass/polyester panels but is not significant.     

Effect of geometry on compressive failure of notched composites

The compressive failure of carbon fibre-epoxy laminates is investigated theoretically and experimentally. Panels with a single edge notch, a central notch or a central hole are considered. The failure mechanism is by microbuckling in the 0° plies and is accompanied by delamination and plastic deformation in the off-axis plies [1]. To predict the critical length of the microbuckle and the failure load, the microbuckle is modelled as a cohesive zone. The magnitude of the normal compressive traction across the microbuckle is assumed to decrease linearly with increasing overlap of material on either side of the microbuckle. The relative effect of the specimen size and a bridging length scale is investigated to illustrate the transition between small-scale and large-scale bridging. If the bridging length scale is small compared with the specimen dimensions, the specimen fails when the stress intensity at the notch tip equals a critical compressive stress intensity factorK _IC . When the bridging length scale is not small compared with either the initial defect size or the unnotched ligament length then it is necessary to include the details of the traction across the microbuckle to predict the failure load accurately.     

Interface strength effects on the compressive-flexural/shear failure mode transition of composites subjected to four-point bending

Wang's microbuckling model [1] has been extended to oriented fibre composites loaded in four-point bending. The modified model shows that when any internal parameter (e.g. the interface strength) is changed, the resultant change of the interlaminar shear strength and that of the compressive strength are always of the same sense. Additionally, the degree of change of the shear strength is always larger than that of the compressive strength. As a consequence of this conclusion, a four-point bend test piece which normally fails in the flexural compressive mode may fail in the shear mode upon interface degradation. This was rationalized with the aid of the failure-mode transition diagram [2]. This diagram has been used to explain the change in bend failure mode resulting from a change of the external parameters, such as the span-to-thickness ratio, and the fibre fraction. Experiments were conducted to verify such a failure-mode transition behaviour for fibreglass composites of different interface conditions, when the flexural compressive failure mechanism was of the microbuckling type.     

The mechanical performance and impact behaviour of carbon-fibre reinforced PEEK

The mechanical performance and impact behaviour of carbon-fibre reinforced polyether-ether ketone (PEEK) with a (0, ±45) lay-up has been compared with that of a similar carbon fibre/epoxy laminate. Differences occurred because of the greater shear strength and lower shear modulus of the carbon-fibre reinforced PEEK. When compared with the carbon fibre/epoxy laminate, carbon-fibre reinforced PEEK was more notch sensitive in tension and had a lower undamaged compressive strength. However, after impact, the residual compressive strength was significantly greater for carbon-fibre reinforced PEEK because delamination was less extensive.     

Synthesis and property of short tungsten fibre/Zr-based metallic glass composite

The short tungsten fibre reinforced Zr_41.2Ti_13.8Ni_10.0Cu_12.5Be_22.5 bulk metallic glass composites with macroscopic isotropic mechanical properties were prepared by infiltration and rapid solidification. The diameters of the tungsten fibres are 300, 500 and 700?µm with a length of 1000?µm and the fibre volume fraction in three kinds of composites is between 60 and 65%. The room temperature compressive deformation behaviours of these composites were investigated systematically. The results show that the strength and plasticity of the composites increase with the decrease in the tungsten fibre diameter. The maximum compressive strength and plastic strain of the composite with 300?µm fibre, respectively, reach 3079?MPa and 37%. The fracture modes of all the composites are shear fracture. The superior compressive property of the composites with short tungsten fibre is due to the competition among different fracture modes and the inhibition effect of the interface on the shear band extension.     

Mechanical behavior and failure process during compressive and shear deformation of honeycomb composite at elevated temperatures

The mechanical behavior and failure mechanism of honeycomb composite consisting of Nomex honeycomb core and 2024Al alloy facesheets were investigated. The compressive and shear deformation behaviors of honeycomb composite were analyzed at temperatures ranged 25–300°C. The compressive and shear strengths of honeycomb composite decreased continuously with increasing temperature up to 300°C. The stress-strain curves obtained from the compressive and shear tests showed that the stress increased to a peak value and then decreased rapidly to a steady state value, which is nearly constant up to failure with increasing strain. The compressive deformation behavior of honeycomb composite was progressed by an elastic and plastic buckling of cell walls, debonding fracture at the interfaces of cell walls, and followed by a fracture of resin layer on cell walls. The shear deformation of honeycomb composite was progressed by an elastic shear deformation, plastic shear deformation, fracture of resin layer on cell walls, and followed by debonding fracture at core/facesheet interfaces. The shear strength of honeycomb composite showed strong anisotropy dependent on the loading direction. The shear strength in longitudinal direction was about 1.4 times higher compared to that in transversal direction due to the different thickness of cell walls mainly loaded during the shear deformation.     

Adiabatic shear banding in a thick-walled steel tube

We analyze the initiation and propagation of adiabatic shear bands in a thick-walled 4340 steel tube with a V-notch in the middle. The material is modeled as strain hardening, strain-rate hardening and thermal softening. The deformations are assumed to be locally adiabatic and the effect of inertia forces is considered. Two different loadings, i.e., torsional, and combined torsional and axial pressure are considered. In each case, the load generally increases linearly from zero to the final value, is kept steady there for some time, then decreases to zero and is kept at zero; thus a finite amount of energy is input into the body. For the combined loading, the magnitude of the torsional loading pulse is kept fixed and the effect of varying the magnitude of the axial pressure preload is investigated. A shear band first initiates in the element adjoining the notch tip and propagates radially inwards. By recording the time when a shear band initiates at the centroids of different elements we determine its speed of propagation in the radial direction to vary from approximately 50 m/s at the instant of its initiation in an element abutting the notch tip, to nearly 90 m/s by the time it reaches the innermost surface of the tube; the speed also depends upon the overall loading rate, and whether or not the loading is multiaxial. The drop in the torque required to twist the tube at the initiation of a shear band is not as sharp as that in a thin-walled steel tube. We compute the distance through which a shear band propagates as a function of the energy input into the body and thus ascertain the energy required to drive a shear band through a unit distance. We also study torsional deformations of a thick-walled CR-300 steel tube, model its thermal softening by a relation proposed by Zhou et al. and use material properties derived from their data. In this case, the speed of a shear band initiating from an element abutting the notch tip is found to vary between 750 m/s and 1,000 m/s at different points on a radial line through the notch tip; this agrees with that observed by Zhou et al. in their experiments on single-notched plates.     

Compressive response of Kevlar-epoxy composites: experimental verification

The initial misalignment of Kevlar fibres in Kevlar-epoxy composites is quantitatively investigated. This misalignment has been found to be one of the most important factors for determining the compressive response of these composites. A theoretical model, which considers initial fibre misalignment and assumes that the compressive response of Kevlar-epoxy composites is dominated by kink band failure, is in good agreement with experimental results. In addition, photomicrographs of the failure surfaces suggest that kink band formation is the predominant failure mode in this composite system.