JOURNAL OF MATERIALS SCIENCE 15 (1980) 677-680 Craze and yield zones in the fracture of rigid PVC H. R. BROWN, TAN HEACHIN* Department of Materials Engineering, Monash University, Clayton, Victoria 3168, A ustralia Two techniques have been used to examine the reported discrepancy between the measured fracture toughness of rigid PVC and that calculated from its crack tip craze. The use of a generalized model of a cohesive zone produced no better agreement between calculation and experiment than the Dugdale model used previously. The discrepancy was probably caused by the coexistence with the craze of a yield zone which was observed by optical and scanning electron microscopy. 1. Introduction The brittle failure of a number of amorphous glassy polymers, in particular PMMA, polystyrene, polycarbonate and PVC, has been shown to occur by a crazing mechanism [1, 2]. PVC is known to show also a non-crazing failure mode [3, 4]. The shape of crack tip crazes has been examined for PMMA and shown to be an accurate fit to the Dugdale model [5, 6] and the stress intensity factors from the crack tip craze displacements using the Dugdale model also agree well with experimental values. This is not the case for crack tip crazes in PVC however; both Mills and Walker [2] and also Brown and Stevens [4] found that in this material the crack tip crazes were much more wedge shaped than would be predicted by the Dugdale model, and also that the stress intensity factors calculated from the craze parameters using the Dugdale model were less than half of those observed experimentally. Mills and Walker suggested that this discrepancy might be caused by the existence of plastic yielding above and below the craze whereas Brown and Stevens thought the use of the Dugdale model for the wedge shaped crazes might be the problem. A fairly similar discrepancy has been observed by Fraser and Ward [7] in polycarbonate. In this case shear lips were plainly visible on the sides of the specimen and they analysed their data on the assumption that the failure energy of the shear lips was higher than that of the craze. Kambour et al. [8] have however noted the *Present address: Esso Australia Ltd, Sale, Victoria 3850, Australia. 0022--2461/80/030677--4502.40/0 9 1980 Chapman and Hall Ltd. existence of plane strain plastic zones as well as crazes in low speed failure in polycarbonate at room temperature. The purpose of the current work is to attempt to discover which, if any, of these suggestions is correct for PVC. This has been approached in two ways, firstly by optical and scanning electron microscopic observation of fracture in a commer- cial grade of clear rigid PVC and secondly by recalculation of stress intensities from crack tip crazes by use of Smith's [9] generalized model of the "cohesive zone". Smith has proposed a generalized model for a cohesive or yielded zone at a crack tip in which the stress across the zone is not assumed to be constant. Using this model it is possible to calculate stress intensity factors from craze shapes with no assumptions about the magnitude of the stress across the craze and so find directly whether the failure energy is all going into this crack tip craze. 2. Theory Smith assumed the existence of an infinite elastic body which contains a centre slit of length 2c on the axis, y = 0, the solid experiencing a remote stress pyy-----O" 0 such that it deformed in plane strain. A small cohesive zone of length R (~ e) was assumed to exist near the crack tip and the co-ordinate system was chosen such that x = 0 at the tip of the cohesive zone. He showed that the stress p(s) across the cohesive zone at a pos- ition x = s is given by 677