Fibril failure, crazing

This raises the interesting question of under what circumstances the craze breaks leading to a system crack/craze which is usually hard to resolve from a crack. By the aid of novel methods, such as acoustic emission, fibril failure can be detected, however, this problem has yet not been solved satisfactorily. 

In the previous sections properties of crack tip crazes in thermoplastics within two different regimes of damage behavior have been described, that of stationary and slowly propagating cracks. During steady state slow crack propagation, as described above, at a particular crack speed the crack tip is preceded by a craze zone of constant size, indicating an equilibrium between fibril formation and fibril failure as demonstrated exemplarily in Fig. 3.27 a by two interference micrographs of the... 

On the other hand, under subcritical load of long duration, when the crack does not grow, an increasing craze size (see Sect. 3.1) indicates that bulk material is fibrillated without any detectable synchronous fibril failure as shown by the interference optical micrographs in Fig. 3.27b taken in a time interval of 3.4 10 s during... 

Examination of the fracture surfaces of PESO specimens in Figure 11 reveals that the fracture is characteri.sed by the craze fibrils failure regardless of the specimen thickness and loading rate. This was quite different from the fracture characteristics examined in Figure S when the LBA ratio was varied. 

An interesting aspect in the understanding of the micromechanics of the interface is the fibril failure mechanism in the crazing regime. As pointed out in the previous section, two mechanisms are possible chain scission or chain dis-... 

In neat polymers the fatigue failure process is quite similar, but in certain amorphous thermoplastics, such as polystyrene, phase crazes are formed during the initiation [13] with the subsequent fatigue crack propagation phase to final failure. (Crazes resemble regions with many cracks bridged by oriented fibrils). 

It is likely that both occur depending on the conditions and morphology of the material. The first step in considering the likelihood of fibril failure is to look at the stress levels within the fibril. Because of the reduced cross-sectional area in the craze region, the true stress on the fibrils (uct) can be considerably higher than the stress at the craze boundary (draw ratio of the fibril (A.) ... 

This approach incorporates the stress-concentrating effect of cross-tie fibrils, widely observed in crazes in glassy polymers (compare Figure 14.14). In the absence of any stress-concentrating effect, that is, for a —> 0, a time-independent fibril failure criterion oy implies crack advance can never occur, because the stress in a given fibril can never exceed Oc- This result has been confirmed by more detailed micromechanical modeling, and is important in that it provides a direct link between the... 

The aim of the model is to find the relation between the energy dissipated in growing the crack tip craze, as that is the macroscopic toughness and the local force at the fibril closest to the crack tip which controls the molecular level failure... 

The fracture energy cannot be related to the failure of chemical bonds which may contribute only with a few Jm-2. Furthermore, the possibility of crazing is not allowed in thermosets because fibrils cannot exist due to the high crosslink density. So, in the case of high-Tg cross-linked materials the main source of energy absorption before failure is the yielding of the network. This assumption is obviously valid only above the ductile-brittle transition temperature (Fig. 12.5), where yielding is temperature-dependent. ... 

The craze is modeled by an elastic anisotropic medium with Young s moduli i and 2 corresponding to the stiffness of the cross-tie and the main fibrils ( 1 2). As the strip thickens along the direction 2 (Fig. 4b), the elastic energy density W is stored so that the energy release rate at failure... 

---