MUHAMMAD SULIMAN

Deformation characteristics in K0 consolidated soils considering ground rebound behavior

Hirofumi TOYOTA

Although growing infrastructure requires vast area for the construction, surface land availability is limited. Therefore, construction sites have progressively shifted to deeper underground, where large-scale excavations generate significant stress relief at the base of the excavation pit, resulting in upward ground movement commonly referred to as rebound or heave. Accurate prediction of this phenomenon has still difficulty that is considered as one of the persistent challenges in geotechnical engineering practice. Conventionally, rebound magnitudes have been estimated using empirical methods or laboratory consolidation tests, and cyclic loading tests to derive deformation coefficients. However, these approaches inadequately capture stress relief behavior inherently associated with excavation-induced rebound, particularly under extensional stress conditions that characterize the unloading stress path in the field.

This study establishes a more effective experimental framework for estimating rebound deformation by investigating the secant shear modulus (Gsec) of K0-consolidated soils under compressional and extensional loading paths. Triaxial tests incorporating bender element (BE) and local small strain (LSS) tests were conducted on undisturbed and reconstituted specimens from a metro construction site. Two layers were mainly used for discussion, depth of which are 8.50?9.50 m and 21.50?22.50 m, classified as silty clay. The results of other two layers, which are depth of 13.50?16:50 m (silty sand) and 30.5?33.50 m (silty clay), were also briefly summarized. The undisturbed specimens exhibited high stiffness and a large void ratio, reflecting strong interparticle cementation and intact natural soil fabric. The Gsec was normalized by the initial shear modulus (G0) for direct comparison considering soil types and consolidation conditions.

The normalized results revealed a pronounced directional dependence in stiffness degradation. Under K0-consolidated undisturbed conditions, extensional loading retained higher normalized stiffness to larger shear strains than compressional loading. Furthermore, degradation trends of undisturbed soils closely resembled those of reconstituted K0-consolidated specimens, supporting that reconstituted samples can be used as a practical alternative of undisturbed samples. However, isotropically consolidated reconstituted specimens failed to reproduce normalized degradation of K0 consolidated undisturbed samples, which most closely represents field rebound conditions. On the other hand, isotropically consolidated reconstituted specimens showed consistent normalized trends with K0-consolidated undisturbed specimens under compressional loading.

These findings provide possible low-cost testing methods, an alternative to the most reliable method using K0-consolidated undisturbed specimens.