Experimental and discrete element modeling studies of the trapdoor problem: influence of the macro-mechanical frictional parameters

Abstract

Granular soils have the inherent ability to develop load transfers in their mass. Mechanisms of load transfers are used as a basic principle of many civil and geotechnical engineering applications. However, their complexity makes it difficult to formulate relevant design methods for such works. The trapdoor problem is one of the ways to reproduce load transfers by the arching effect in a granular layer in non-complex conditions. In addition, many analytical solutions for the prediction of load transfer mechanisms are based on the trapdoor problem. However, some of the parameters required are still being widely discussed, in particular the ratio of horizontal stress to vertical stress. For this paper, an experimental device for trapdoor tests in plane strain conditions was created and several geomaterials were tested. Three phases in the response of the materials were consistently observed. Each of these phases corresponded to a specific displacement of the trapdoor. A first phase of high load transfer was observed followed by a transition phase which was followed by a critical phase for which the load transfer amplitude increased and stabilized. Analytical solutions and experimental values of load transfers were compared. Considerable differences between the stress ratio needed to fit the experimental data and the stress ratio proposed in the analytical models were noted. Based on the conclusions of the experimental study, the discrete element method was used to model the same trapdoor problem. A wide range of granular materials was modeled and tested in the trapdoor problem. The three phases in the response of the layer were also observed in the numerical modeling. In addition, it was shown that the shear strength of the material is the key parameter of load transfers: peak shear resistance for the small displacements of the trapdoor and critical shear strength for the larger displacements. A micro-mechanical analysis showed that the effective stress ratio in the sheared zone does not vary as much with shear strength. Stress ratios here were again greater than those proposed in the analytical solutions. Nevertheless, the relevance of the solution of Terzaghi was confirmed as soon as the stress ratio was correctly chosen.