Modeling results summarized above confirm that kink-bands are strain localization bands and as such their initiation is necessarily controlled by the rheology of the material, stress state and deformation conditions. Once initiated, subsequent evolution of kink-bands is largely kinematic. Kink-band boundaries migrate through the material, rotate to unfavorable orientations and cease to operate as a deformation mechanism as strain increases. New kink bands are continually developed. To develop kink-bands, the material must not be too anisotropic for high degree of anisotropy will favor pervasive kinking at the expense of localized kink-bands. Kink-bands and kinks are products of competing mechanisms (Figure 14). The latter do not result from widening of the former.
The deformation mechanisms map can be readily applied to natural kink-bands. Friction angles of common rock types are experimentally constrained [e.g., Carmichael 1989]. By examining the geometry of kink-bands and kinks in an area, one can use Figure 14 to estimate of the degree of anisotropy of the rock.
An anisotropic rock is often more complicated than a ubiquitous joint body. First the ubiquitous joint model has an axial symmetry because in the plane of the ubiquitous joint there is no rheological difference. A rock may, however, have a linear fabric in addition to a foliation and therefore exhibit anisotropy lower than axial. Second, the rheology of a rock may change during the course of deformation. For example, a rock behaving viscously under high-grade metamorphic conditions may be transiently plastic if the fluid pressure is elevated due to partial melting or dehydration metamorphic reaction, leading to significant reduction of the effective confining pressure. If the rock has a planar fabric as well, kink-bands and kink folds may accordingly develop even though the dominant rheology of the rock is in the viscous field.