Extrinsic crazes

For experimental convenience all the scattering curves were taken from samples after stress relief. This should not be a severe limitation. In fact, in subsequent sections experimental evidence will be presented that the fibrillar structure of intrinsic crazes in PC is not strongly affected by stress relief as previously shown for extrinsic crazes by Brown and Kramer. ... 

Evidently, the application of Equation (37) requires to be sufficiently different from Q, i.e. the number of crazes must be large. Extrinsic crazes do, generally, not fulfill this requirement. However, as shown in Figure 2, intrinsic crazes in PC cause a marked decrease in density and, hence Vf may be evaluated from the above Equation. 

The equatorial SAXS curve of extrinsic crazes in PC is shown in Figure 17 for a sample which has been stretched at T = 129 °C to = 1.8, i.e. to 7. < The scattering curve exhibits the well pronounced interfibrillar interference maximum discussed in the previous section. From the position, s, of this maximum the average fibril diameter, D, may be evaluated. If the volume fraction of craze I fibrils is assumed to be vj = 0.5 (see Section 3.2), Equation (49) yields D = 37 nm, which compares well with results obtained under approximately the same drawing conditions by Paredes and Fischer and Brown and Kramer... 

The meridional scattering curve of intrinsic crazes has not been analyzed in detail. However, it may be worth mentioning that this scattering component continuously increase up to very high intensities at small reciprocal vectors. The same type of scattering behavior has previously been reported for extrinsic crazes J. Brown and Kramer assumed that internal reflections at the craze-matrix boundary strongly contribute to the intensity measured along the meridian. 

Vf" remains virtually constant as a function of the pre-orientation. However, as shown in Figure 24, D increases linearly with the total extension ration, k, at craze initiation. It is interesting to note that the fibril diameter of D = 36 nm which is obtained by extrapolating the straight line to = 1 is in excellent agreement with the value of D = 37 nm measured for extrinsic crazes. This result seems to indicate that the different fibril diameter of extrinsic and intrinsic crazes primarily reflects the distinct extension ratios X and k of the matrix at craze initiation. However, further investigations are necessary to substantiate the above result. 

The structural analysis of intrinsic crazes in PC has been carried out by SAXS. The fibrillar microstructure of these crazes gives rise to pronounced scattering effects which enable a detailed analysis of the craze structure in terms of both the voliune fraction of craze fibrils and of the fibril diameter. This analysis showed that the microstructure of intrinsic and extrinsic crazes is largely different. There exists some evidence that the distinct microstructure primarily reflects the different stress-strain states of the matrix at craze initiation. Further investigations are necessary to answer... 

The author is well aware of the fact that many aspects which have been treated in the extensive literature on extrinsic crazing have not been considered in this article and that more information is needed for a comprehensive account of the observed craze phenomenon. For instance the recent work on the intrinsic crazing of PC and on related phenomena which has been re wed here has primarily been based on structural considerations. It is believed that future work on the kinetics of craze formation and on the underlying molecular dynamics of the system may contribute considerably to a more detailed account of this phenomenon. Nevertheless, it is hoped that this work has opened up some new paths which may lead to a better understanding of the phenomenon of cavitational plasticity in polymers. 

Recently, Dettenmaier and Kausch have observed an intrinsic craze phenomenon in bisphenol-A polycarbonate (PC), drawn to high stresses and strains in a temperature region close to the glass transition temperature, T. This type of crazing is not only initiated under extremely well defined conditions which reflect specific intrinsic properties of the polymer but also produces numerous crazes of a very regular fibrillar structure. These crazes were called crazes II in order to distinguish them from the extrinsic type of craze, called craze I. As shown by the schematic representation in Figure 1, a detailed quantitative analysis of intrinsic crazes in terms of craze initiation and microstructure was possible. The basis of this analysis and the results obtained are reviewed in this article. 

Finally, our interest will be limited here exclusively to the phenomenon of crazing in heterogeneous polymers. Thus, apart from the considerations of improving toughness by manipulation of the processes that govern the craze flow stress and, thus, rendering the extrinsic flaws inoperable that result in craze fracture, we will not consider the mechanics of fracture of crazable polymers. A brief survey of this subject related to the crazing process can be found elsewhere... 

As we discussed in Sect. 2.3, in aU crazable polymers toughness is limited by fracture which occurs almost exclusively by the breakdown of craze matter under stress. When no extrinsic imperfections are present, craze fibrils fracture as a result of molecular level stress concentrations when load-bearing molecules fracture and set off an unstable cascade of molecular scissions. When large, micron-si d particulate... 

Homo-PS is known to be brittle in tension at room temperature in unmodified form, as Fig. 13.3 demonstrates. It has a compressive yield strength of around 103 MPa that, with a substantial strength-differential effect, translates into a tensile yield strength of 73 MPa, and undergoes plastic flow if its brittleness can be suppressed. Experimental evidence, such as that in Fig. 13.3, shows that homo-PS undergoes brittle behavior initiated from surface flaws, and that elimination of these is impractical, primarily because, even if that could be achieved, crazes could still be initiated at free surfaces, as is discussed in Chapter 11, and craze matter breaks down from either extrinsic or intrinsic imperfections in craze matter at stress levels of around 40 MPa at 293 K. 

---