Influence of a permanent low-friction boundary on rotation and flow in rigid inclusion/viscous matrix systems from an analogue perspective

Abstract

Natural occurrences and recent experimental work show that a low-friction inclusion/matrix boundary can be responsible for antithetical rotation and development of stable shape preferred orientations in simple shear. The flow of a viscous matrix around a rigid inclusion to which it may or may not be adherent is still not well studied, but it is relevant to the understanding of the behaviour of structural elements in mylonite zones. We used two-dimensional (2-D) analogue experiments to address these issues. The experimental results with a permanent low-friction inclusion/matrix boundary (nonadherent mode) show the following. (1) The rotation behaviour of inclusions in this mode is markedly different from the theoretical predictions and experimental results for the adherent mode. (2) Inclusions with aspect ratio equal to 1 rotate synthetically at a rate that depends upon inclusion shape and orientation. (3) Ellipse-, rectangle- and lozenge-shaped inclusions rotate antithetically when starting with their greatest axis parallel to the shear direction. (4) Back rotation is limited in all cases studied, and the angle between the inclusion greatest axis and the shear direction represents a stable orientation, which depends on inclusion aspect ratio and shape. (5) A metastable orientation exists, which is strongly dependent upon inclusion shape and aspect ratio, and separates fields of antithetic and synthetic rotation. Our experimental results show that the overall flow pattern is bow tie-shaped in adherent and nonadherent modes. However, there are major differences in the way the matrix flows near the inclusion. (i) In the nonadherent mode, the nearby matrix flows past the inclusion, whereas in the adherent mode, the nearby matrix flows with and follows the inclusion. Therefore, in the adherent mode, passive marker lines parallel to the shear direction and streamlines show considerable deflections at the inclusion crests, in marked contrast with their straight character in the nonadherent mode. (ii) Stagnation points or closed flow lines near the inclusion were not observed in the nonadherent mode, which means that there is no closed separatrix around the inclusion in this mode, despite the fact that the overall flow shape is bow tie. (iii) In the adherent mode, the line of flow reversal is stable throughout deformation, but in the nonadherent mode, it changes position and orientation with progressive shearing. This shifting of the line of flow reversal can be an important factor controlling rotation behaviour in the nonadherent mode. (iv) In the nonadherent mode, the inclusion behaviour is similar to that observed in confined flow. (v) The flow pattern in the nonadherent mode provides an explanation for the observed lack of drag folds associated with small-scale rigid inclusions in mylonites.