Literature relating to the fundamental nature of the fibrous state is reviewed with the goal of illustrating evidence for the existence of a possibly fundamental morphological unit, the elemental microfiber or fibril. Following a short summary of the rather well documented situation in natural fibers, the problem of the microfiber in synthetic fibers is probed in terms of phenomena in solution and bulk polymeric systems. The reviewed literature indicates the existence of unique, elemental microfibers in synthetic fibers is largely tenuous; however, there is wide evidence of the existence of complex elongated structures axially aligned within macrofibers which future work may resolve into a central microfiber. The structural complexity appears to simplify for transformations occurring in high shear fields. This review covers articles through 1971, and it is notable that in 1971 approximately 100 articles of high relevance appeared in major journals.

An experimental method initiated by Williamson and Busse has been extended to produce veil‐free shish kebab structures in linear polyethylene without need for severe recovery treatments. The technique involves the dilution of polymer in a concentrated solution of low molecular weight paraffin followed by a simple isothermal shear. The resulting shish kebab structures are isolated from the wax matrix and subjected to electron microscopical and electron diffraction study. From the diffraction patterns, it is shown that the shish‐rich ordered structure is similar to the orthorhombic structure of crystalline linear polyethylene, but with certain significant deviations. The resolvable morphology of the shish appears to be characterized by a length‐to‐diameter ratio which is consistent with a prediction based upon Wulff's theorem utilizing certain relevant surface energy values which have been obtained by others for polymeric crystals.

The load induced microscopic structural changes in highly oriented Nylon filaments were followed by changes in the x-ray fiber diffraction pattern as a function of the resulting strain. Observations made included (1) changes in the dimensions of the fiber spots, and (2) a decrease in ratio of the intensity of amorphous to crystalline x-ray scattering. These new results for Type 200 du Pont Nylon indicate a change in the micro deformation mechanism at a strain corresponding to the inflection of the load-strain diagram (14 percent elongation). The mechanism for smaller plastic strains appears to involve a rotation of the crystallized domains into alignment with the fiber axis, accompanied by a small over-all additional crystallization. For strains larger than the inflectional strain, the mechanism seems to involve a progressive reduction in the domain size, and a reduction in alignment with the fiber axis. Both effects appear to saturate in the vicinity of the rupture strain. These results are discussed in terms of load induced effects found in metal single and polycrystals. It is suggested that the qualitative features observed may be representative of a wide class of load or strain induced micro deformation reactions in the solid state.

An elementary theory of the bending of two, coupled, thin elastic elements is presented as a model for computing the curvature of bicomponent filaments. The curvature is expressed in terms of a differential extension between the two elements at the moment of "activation." The model requires expressions for the deformational energy of the interacting elements, and an explicit "material" relation expressing the details of the interactions between the "driving" and "driven" elements of the filament structure. It is closed by minimizing this energy with respect to a parameter measuring the inefficiency ΔL/L2 of the activation strain ε. Curvatures computed for near-unity ratios of rectangular cross sections and elastic moduli yield results comparable to previously published work for both ratio values near unity (the Timo shenko limit). At larger and smaller values of these ratios the curvatures obtained by the new model are lower than those obtained previously and in limited cases appear to approximate experimental experience. A discussion of the general problem of computing and understanding practical crimp situations is given.

No abstract