Craze length

In the higher temperature range, where a single craze develops at the crack tip, it has been shown [46] that a breaking time r of the craze fibrils under a stress, ac> can be defined from the craze length S and the crack-craze velocity vc by ... 

Parameters in the craze initiation criterion of Sternstein et al. [24] Parameters in the craze initiation criterion of Oxborough and Bowden [26] Critical craze thickness Craze length... 

Depending on the material, a critical molecular weight for the observation of a stable craze has been found, for PMMA [29,30] and for PC [42], Below this critical value, crazes are not seen by interferometry and the material is very brittle. The molecular weight has to be sufficiently large (about Mw = 3 x 105 g/mol for PMMA and Mw = 12 x 103 g/mol for PC) for the development of a stable craze. The critical craze thickness and craze length (Acr and Ac) are also temperature dependent [29,30,43,44] and this effect is amplified with increasing molecular weight [29,30]. 

The similarity between the measured craze zone and the calculated Dugdale plastic zone is apparent when Figs. 3 and 7 are compared. Before applying the Dugdale formula to the measured craze zone some points of detail should be noted. Whilst the position of the crack tip is known fairly precisely, the location of the craze tip and hence the craze length s cannot be measured directly, but can — in a similar way to... 

Fig. It a and b. Variation with crack speed of the dimensions of the craze zone at the tip of a propagating crack in PMMA a maximum craze width 2Vj b craze length s...

Figure 11 a shows the maximum craze width 2v as a function of crack speed d and Fig. lib the corresponding craze length s. Initially the results were interpreted as... 

An example of the growth behavior of crazes in a liquid environment is shown in Fig. 20 which is taken from the results of Williams and Marshall They measured the craze length versus loading time at different Kj-values in PMMA specimens immersed in methanol. The time-depettdent craze-behavior was interpreted in terms of a plasticisation mechanism incorporating the effect of the fluid Due to its porous nature the craze has a very high area to volume ratio so that penetration of the fluid by only a small distance leads to a complete plasticisation of the fibrils and a subsequent drop in the load carrying capacity (t of the fibrils the material effectively behaves as one with a lower craze stress oojc (a < 1). 

In all papers mentioned above the characterization of the craze growth has been performed by measuring the craze length. By application of the optical interference method the essential information on the growth of craze width is awiilable. Figure 22 shows the interference fringe patterns and measured craze zone sizes at a stationary crack tip in PMMA loaded in a creep test at a constant K,-value at loading times of 150 s and 2 1(F s... 

Fig. 23a and b. Growth of a primary craze at a stationary crack tip in PMMA in air, K, = 19 N/mm a craze length s as a function of loading time t b maximum craze width 2v as a function of loading time t... 

The craze length s in high molecular weight PMMA is shown in Fig. 34 as a function of AKi- This curve is also based on about 350 individual measurements. The craze length s increases linearly with AKi, from about 10 pm up to about that value measured for a continuously propagating crack, i.e. s = 35 3 pm An increase of s with AKi can also be deduced from the data on craze lengths in vinylurethane and PC measured in the relaxed craze state as is shown in Fig. 35. It is interesting... 

Fig. 35. Increase in craze length s with AK, in PC 108) Vinylurethane > measured in the unloaded state...

In order to obtain information on the craze growth and on the fibrillation process the maximum craze widths are plotted against craze lengths s for high molecular... 

The quantitative development of craze length s and maximum craze width 2v at the crack tip are given in Fig. 50 as functions of the cycle number N for several successive crack jumps. In this experiment (AK, = 15.2 N/mm, R < 0.1) the crack remained stationary up to a critical cycle number Ne which is connected with a critical end length Se or end width 2v respectively. Together with the result of Fig. 49 it is obvious that the crack does not go through the craze all the way up to the craze tip as is usually assumed It stops well ahead of the craze tip, even at this... 

Fig. 50a and b. Craze growth between successive crack jumpsinPVCatAKj= 15.2N per a Craze length... 

The influence of AK, on craze length s in PVC is shown in Fig. 53. Both characteristic craze lengths, s just before and just after a crack jump, show a linear increase with AK, thus leading to a similar result as that observed during continuous fatigue crack growth (see Figs. 34 and 35). 

Fig. 53. Influence of AK, on craze lengths s and band width b in PVC during discontinuous crack growth...

The rate of craze front inactivation in this geometry will be distinctly dependent on mutual arrest of craze fronts when they encoimter within a critical distance 8, permitting the craze fronts to sweep out a mean free path of 1/2 in each direction from the source particle at the center, as shown in Fig. 3. Considering that the typical craze front per cell of P5 will be on the average only half as long as the periphery of the cell, we have for the active craze length per unit volume... 

Many craze parameters are molecular weight dependent. As M rises, craze length, craze width and craze density all continue to increase. In PS, the craze fibril diameter decreases slightly with increase of M but it increases with cycles of loading. 

---