Fractography of highly crosslinked polymers

Abstract

The fracture surfaces of highly crosslinked polymers sometimes include regions in which regular arrays of features are disposed along tracks running in the direction of crack propagation. Generally the features appear to be fibres but in some cases there is evidence that they are curls of rolled-up film. The features may remain attached to the tracks, be partially detached, or entirely missing. Such effects are illustrated by reproduction of results for a phenol-formaldehyde resin in Fig. 1 [1, 2]. Subsequently, similar effects were observed in other highly crosslinked polymers [3] such as polyesters [4], epoxy resins [5] and poly(ethylene glycol dimethacrylate) [6]. The purpose of the present letter is to suggest adoption of a mechanism proposed long ago by Preston to account for "stries" formation in silicate glasses, igneous rocks, and bitumen [7]. This mechanism dealt with the consequences of fracture in an ideal brittle material but, in order to account for the cylindrical shapes of the features actually observed in organic networks, it is necessary to invoke, additionally, the occurrence of localized plastic deformation. Preston's analysis made repeated use of the acceptable premise that crack propagation always proceeds in a direction at right angles to the principle tensile stress in the immediate neighbourhood of the crack tip. He then accounted for "stries" formation by analysis of the consequences of a sudden change in orientation of the principle tensile stress. This was illustrated by reference to a continuous crack front which may be imagined as moving upwards through a glass block (Fig. 2a). Suppose now that the front suddenly moves into an uncracked region in which the tensile stress is at a slightly different orientation. The entire crack