[1] estimated the applied stress at which matrix cracking takes place with the fibres left completely intact. In Marshall and Cox’s work [2], a shear lag analysis, i.e., a constant shear frictional stress, is used to deal with interfacial sliding resistance. This is valid for a composite system with a weak or unbonded sliding interface. Also, different forms of bridging laws have been presented in the past decade to account for the effects of fibre debonding and fibre sliding [4,5]. Further, the effect of interface roughness on the bridging law was analysed by Liu, Zhou and Mai [6]. Even so, the effect of fibrematrix interface debonding and sliding on the fracture process has not yet been thoroughly investigated. In low-density fibre composites a crack is bridged by a few fibres, and the smeared-out model cannot be usefully applied without loss of accuracy. Hence, there is a need to understand the bridging effect by discrete fibres over a finite length. Meda and Steif [4] discussed several reasons for poor agreement between experimental observations and theoretical predictions based on the continuous distribution of tractions along the crack-faces. It is desirable to develop an independent model where discrete fibres are analysed explicitly, rather than implicitly as in earlier crack-bridging models. Fibre breaks play an important role in the softening behaviour of fibre-reinforced composites. However, most solutions available for crack resistance (R)-curves are obtained from approximate shear lag analysis and smeared-out representation of the bridged area and do not include fibre breakage. Also, fracture of composites involves more than one physical process. Thus, rupture in unidirectional fibre-reinforced composites is associated with brittle matrix fracture, fibre pullout and fibre fracture [7]. In principle, distinct failure mechanisms cannot be incorporated in a unified cohesive zone model, although attempts have been made to do so. Hence, matrix cracking and fibre bridging should be modelled separately by a two-part description. That is, the coexistence of a fibre bridging zone and a matrix cohesive zone, whose stresses depend on the crack-face opening displacement. The aim of this work is to study the toughening mechanisms of aligned fibre composites by finite element analysis. Two failure mechanisms, fibre-bridging and matrix cracking, are considered separate entities, rather than by a single smeared-out bridging law. A bi-linear constitutive relation is employed to describe the fibre-matrix debonding and frictional sliding process. A softening cohesive law is used for matrix separation to avoid the stress singularity in conventional bridged-crack models. Numerical results are calculated for a carbon fibre-epoxy unidirectional composite with a low fibre density. The effects of the interface properties and fibre breakage on the fracture toughness of the composite are presented and discussed in detail. THEORETICAL ANALYSIS Bridging Law during Fibre Pullout Consider an infinite uniformly aligned unidirectional fibre composite containing a semi-infinite crack. The crack-face separation and crack growth are restrained by the bridging fibres. Descriptions of the fibre bridging stress and crack-face displacement relationship can be obtained by a theoretical analysis of a fibre pullout process [8] as typified by the stress-displacement curve in Fig.1, in which σ is applied stress at fibre end and δ is pullout displacement. σ1f and f 1δ are the initial debonding stress and displacement, respectively. Two parameters, σ1f and the slope κ (see Fig.1) are used as indicative measures of the interfacial bond strength and friction coefficient at the partially debonded interface, respectively. They define the bilinear law of the stress-displacement relation prior to the full fibre pullout from the matrix (which is not shown). Computer simulation [9] study shows that a strong bond gives a high initial debonding stress and a high interfacial friction. However, a very strong bond will increase the full debonding stress but decrease the pullout-displacement due to fiber breaks.
RkJQdWJsaXNoZXIy MjM0NDE=