Craze microstructure

A TEM micrograph of a thick region of a craze (near its origin) is shown in Fig. 8 a. While one can detect the fibrillar nature of the craze, the strongly 

Nevertheless one can idealize the fibril structure as a forest of isolated, cylinderical fibrils, all oriented perpendicular to the craze surfaces. Such an array of fibrils in the 5 to 10 nm size range should scatter X-rays strongly in the small angle scattering regime One expects to see the scattering as a disk in reciprocal space 

An alternative model is shown schematically in Fig. 10. Just as for the craze tip, the strain-softened polymer being actively deformed at the bulk polymer interface is idealized as a thin layer of non-Newtonion fluid. The velocity v of plastic advance of the craze interface depends on the gradient in hydrostatic tension Van as 

Following the treatment of Fields and Ashby for (ao) we assume it to be proportional to the average tensile stress S on the craze interface 

It seems reasonable to assume that the fibril structure (value of Dq) with the fastest growing interface will be that ultimately observed in turn from Eqs. 15 and 19, 

The most extensive and definitive experiments on craze microstructure in homopolymers were carried out by Kramer and co-workers (Kramer 1983 Kramer and Berger 1990). In these experiments crazes were formed in thin, electron-transparent cast films of polymer deposited on copper TEM grids and firmly bonded to them by solvent vapors. When such films were subsequently stretched slightly, crazes were formed in the stretched films, permitting in-situ study of their microstructure and its mechanical response in the elastically stretched state, under tension, eliminating any possibility of relaxation in craze matter that might have occurred in unloaded fibrils. 

Specific SAXS experiments of Paredes and Fischer (1979) showed that for PC the product of the fibril diameter D and the flow stress Oy (the tensile plastic resistance) remains constant, irrespective of temperature, as shown in Fig. 11.10, at a value of 0.54 J/m that was labeled as the fibrillation energy of the polymer by them and was later demonstrated to be the base surface energy of the polymer 

From their studies and those of others, Kramer and co-workers concluded that the drawing of craze fibrils is basically a nano-scale variant of the plastic drawing of fibers and bars of polymer considered in Chapter 10 (Kramer 1983). This drawing behavior is that of an entangled polymer network in which the 

---

As the craze microstructure is intrinsically discrete rather than continuous, the connection between the variables in the cohesive surface model and molecular characteristics, such as molecular weight, entanglement density or, in more general terms, molecular mobility, is expected to emerge from discrete analyses like the spring network model in [52,53] or from molecular dynamics as in [49,50]. Such a connection is currently under development between the critical craze thickness and the characteristics of the fibril structure, and similar developments are expected for the description of the craze kinetics on the basis of molecular dynamics calculations. 

Fig. 14a—c. Craze microstructure in a a thick film (1.2 nm), b a moderately thick film (0.45 lun) and c a very thin film (0.1 tun). From Ref.courtesy Chapman and Hall... 

Craze Microstructure as Revealed by Small-Angle X-Ray Scattering. 84... 

Normally the craze microstructure is not directly visible in the scanning electron microscope. Thus, an etching procedure using oxygen ions was employed to remove the plastically deformed layer at the sample surface. Control measurements on uncrazed samples showed that this procedure does not lead to artifacts. The surfaces were coated with gold to reduce surface charging. 

SAXS has proved to be a very powerful tool for a quantitative analysis of the craze microstructure. This is not surprising since characteristic craze parameters such as the fibril diameters and the interfibrillar spacings frequently fall into the range of 1-5 X 10 nm covered by SAXS. The general theory of SAXS has extensively been treated in the literature (see e.g. Some basic elements of the SAXS theory,... 

In the above Equations x the total area created by the intersection of a plane perpendicular to the cylinder axis and passing through the origin. denotes the interfacial length of the two dimensional cells in this plane. Crazes may be modeled by a system of parallel cylinders. For a realistic description of the craze microstructure, a distribution in fibril diameter, D, must be taken into account. This yields... 

The objective of the following part of this section is to assess the effect of different parameters on the craze microstructure. In these studies the SAXS curve of each sample has been subjected to the data analysis outlined above. 

The craze microstructure of samples stretched to different extension ratios X > X" has been investigated in Figure 19. It may easily be shown that the observed decrease of D with X cannot be attributed to fibril creep. Otherwise, the relation... 

To discuss the effect of other parameters on the craze microstructure an arbitrary reference state with respect to X must be defined. This has been done by taking X = 1.05).". 

The above results suggest the usefulness of a systematic analysis of the craze microstructure by means of light-scattering. In fact, this method could provide valuable information which is not available by SAXS. For instance, the optical anisotropy of the craze fibrils may be derived from the depolarized light-scattering component. 

In Eq. (10) that gives the toughness in dilatational plasticity the factors C and A are dependent on craze microstructure and will not vary significantly. The stress and temperature dependence of the craze velocity while quite determinate in the interface convolution process of craze matter production will also be quite sensitive to micro-structural detail of phase distribution in block copolymers. The appUed stress = Y ... 

Studies of craze microstructure and the surrounding displacements of crazes have established that the only parts around a craze that undergo plastic deformation are concentrated into a process zone at the tip of the craze, and into a fringing layer all around the entire craze body. In the process zone craze matter is generated by one of the two processes discussed above, and fibrils are necked down to the final extension ratio. In the fringing layer, additions are made to craze fibrils by drawing polymer out of half space. Outside the idetifiable parts of a craze, the solid polymer remains entirely elastic while inside the craze body the fully drawn fibers carry the required craze tractions purely elastically in their orientation hardened state at the... 

Fig. 14a and b. Crazes in KRO-1 Resin a general distribution of well defined crazes in the KRO-1 microstructure b detailed view of craze microstructure consisting of drawn tufts of PS incorporating some drawn PB. (From Argon et al. courtesy of J. Wiley and Sons)... 

Figure 20 shows a typical craze microstructure in these strained di-block copolymers, that for PS 600/PB 256. Examination revealed that the scale of the craze micro-... 

Fig. la. Brighl-field transmission electron micrograph (TEM) of typical craze microstructure in polytstyrenc-acrylonitriie) PSAN- b Low angle electron diffraction pattern from the fibrils of the craze in a. Note the main-fibril axis lies primarily along s,. the tensile axis direction and the direction norma] to the craze-bulk polymer interface... 

Suppose now that there is a wide spectrum of craze microstructures with different values of the fibril spacing D,. For the crazes with very small and very large Dq the craze interface velocity from Eq. (9) is miniscule, since for these crazes Vctq is small. Clearly there will be a value of D which maximizes Wq, and hence v, and this value is given by... 

To examine craze microstructure, and to study the effect of molecular variables on craze morphology, the method described by Kramer was followed. Samples of polymers were cast in the form of thin films, strained in tension while bonded to carbon-coated grids, and examined in the transmission electron microscope either before or after staining. The TEM observations were made with an Hitachi HU-11 A unit or with a JEOL JEM-IOOCX unit, operating usually at 75-80 kV. Fracture surfaces of many bulk samples were coated with a thin layer of gold-palladium and examined by an Etec scanning electron microscope. 

---