Adhesional pressure as a criterion for interfacial failure in fibrous microcomposites and its determination using a microbond test

Typical interfacial strength parameters calculated from the microbond and pull-out tests are the ultimate (local) interfacial shear strength, tau(d), and the critical energy release rate, G(ic). These two parameters are often considered as criteria for interfacial failure. Several years ago, we proposed a new interfacial parameter, adhesional pressure (sigma(d)), which is the normal (tensile) stress component at the fiber-matrix interface at the moment of the debonding onset near the crack tip. Adhesional pressure has the following advantages as an interfacial parameter and eventual failure criterion: (a) it corresponds to the actual mechanism of crack initiation in the microbond test (interfacial debonding starts in tensile mode), and (b) the sigma(d) value is directly proportional to the work of adhesion, W-A, between the fiber and matrix surfaces. In this paper, we investigated, both theoretically and experimentally, the applicability of the adhesional pressure as an interfacial failure criterion. Several fiber-polymer pairs have been tested using a microbond technique, and local interfacial parameters (tau(d), G(ic) and sigma(d)) for these systems have been measured. Assuming sigma(d) = const, we analyzed the initiation of interfacial crack (the debond force, F-d, as a function of the embedded length, l(e)) as well as crack propagation (variation of the crack length with the load applied to the fiber) and dependencies of the peak force, F-max, and the apparent shear strength, tau(app), on the embedded length. Residual thermal stresses and interfacial friction were included in our analysis. A comparison with two other interfacial failure criteria (tau(d) and G(ic)) was made depending on specimen geometry (cylindrical specimens; spherical matrix droplets). It was found that all three parameters satisfactorily describe interfacial failure in a microbond test, yielding asymptotically (at large embedded lengths) very close predictions for F-max and tau(app) as functions of the embedded length. They also predict equally well the interfacial crack growth under increasing external load for not very large crack lengths. For shorter embedded lengths, the three interfacial criteria yield substantially different results, e.g. finite tau(app) value as l(e) -> 0 from the tau(d) criterion, infinite tau(app) from G(ic) = const, and zero tau(app) from sigma(d) = const. In this range of l(e)'s, the local interfacial strength, tau(d), appears to be the best failure criterion. (c) 2006 Elsevier Ltd. All rights reserved.