Characterizing internal flow mechanisms in reverse extrusion tests using X-ray CT and marker tracking

Abstract

The reverse extrusion (RE) test has been used to evaluate the undrained shear strength of fine-grained soils, but its interpretation is complicated by uncertainties in failure mechanisms. This study adopts X-ray computed tomography (CT) combined with a marker-based tracking method to characterize internal flow patterns during RE testing of kaolinite samples prepared at different water contents. Fine sand particles embedded within the samples served as tracking markers, enabling detailed characterization of internal deformation through sequential X-ray CT scans and advanced image processing techniques. An artificial neural network (ANN) model was developed to enhance marker linkage between scans, using optimized searching zones based on the axisymmetric displacement fields. The analysis revealed two distinct deformation regions for the tested samples, largely independent of soil water content. Near the loading ram, significant radial inward soil displacements created a dome-shaped shear failure surface, driving soil extrusion to occur via the central orifice. In contrast, the region nearest the closed end of the sample chamber exhibited minimal soil displacement, being somewhat influenced by sidewall friction, entrained air bubbles, and stress transfer from the extrusion zone. The observed deformation patterns, characterized by arched shear surfaces and stress rotation zones, differ significantly from previously assumed RE models. These findings highlight the need to refine RE test interpretations, as applied to fine-grained soils, in order to account for complex internal flow mechanisms, chamber sidewall friction effects, and the presence of entrained air bubbles, which would result in improved reliability of RE undrained shear strength measurements.