Revisiting strain localization from the view of structured inhomogeneity
Year: 2025
Type: article
Abstract: Strain localization in sandy geomaterials, particularly shear band formation, is influenced by loading conditions, boundary constraints, constitutive behavior, and initial inhomogeneity. Among these factors, the effect of inhomogeneity has often been inadequately represented in numerical studies, either through unstructured randomness or insufficient quantitative analysis across multiple realizations. This study revisits the problem by combining a systematic review of existing literature with extended simulations that emphasize the role of structured state fields. A conditional random field method is employed to generate spatially correlated void ratio distributions that reflect material heterogeneity. Simulations conducted under idealized boundary conditions and using a simple, void ratio-dependent constitutive model show that structured inhomogeneity alone can initiate and guide shear band development in ways consistent with experimental observations. Analysis from a dilatancy perspective reveals how spatial variation in void ratio governs the localization of volumetric strain and influences shear band’s evolution. To further interpret micromechanical mechanisms, two simplified modeling frameworks are introduced: one based on void ratio-dependent elasticity and another representing a two-element structural idealization. These approaches help clarify how micro-scale heterogeneity gives rise to macroscopic deformation patterns, even without advanced constitutive formulations. This study emphasizes that an accurate representation of initial state inhomogeneity is crucial for understanding strain localization and the resulting macroscopic behavior. Accordingly, this factor should be explicitly accounted for in both experimental interpretation and the development and calibration of constitutive models.
Source: Computers and Geotechnics
Authors: Zhihui Wang, Roberto Cudmani
Institution: Technical University of Munich
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,640
OpenAlex ID: W4415200622