Model Tests and Numerical Analysis of Subsidence of Reinforced Soil Over a Cavity

Geosynthetics have been commonly used as reinforcement layers to bridge over underground cavities or sinkholes to support upper soil mass. In such applications, the geosynthetics, acting as tensioned membranes, plus the effect of arching in soil, maintain the stability and mitigate the subsidence of the overlying soil mass. This thesis encompasses an experimental study and a numerical analysis to investigate the subsidence of the soil mass induced by the permissible deformation of the geosynthetic reinforcement. The experimental study was performed in a 875×650×50 mm3 (Length × Height × Width) container with 7 trapdoors (125 mm each) at the bottom. One of the trapdoors can be lowered to induce differential settlements. The soil was simulated by cylindrical aluminum bars that have different sizes in diameter but uniform in length of 50 mm. The aluminum bars were intentionally used as the substitutes of granular soil particles to imitate the two-dimensional (2D) situations. A layer of needle punched non-woven geotextile was placed underneath the "soil fill" to serve as the reinforcement. During the tests, the trapdoor was lowered and the induced movements of the "soil particles" and the deformation of the geosynthetic layer as well as the movements of the soil particles were recorded and quantified by photogrammetry methods. The subsidence profiles at different elevations were developed. Furthermore, a numerical analysis using Discrete Element Method (DEM), PFC2D, was performed to extend the scope of the experimental study by varying parameters such as friction angle, cohesion, height of overlying soil mass, and particle size distribution.

It was found that subsidence was decreased hyperbolically in the vertical direction and the lateral influence range appeared to be bounded by a nonlinear smooth curve, which have been simplified into two linear portions: (1) an active soil zone within a plane of 45&deg+&phi/2 near ground surface, and (2) a failure plane of a repose angle. Friction angle shows significantly influence on the soil arching. The normalized settlements were significantly reduced with higher friction angle. With the increase of the backfill height, the maximum deformation becomes greater and the lateral influence range is larger. Cohesion plays an important role in the arching effect. With higher cohesion, the subsidence was reduced significantly. The size ratio shows some influence on the settlement. Basically, with more large particles the subsidence becomes less. Such influence may be an outcome of the friction angle. The mixture with more large particles shows a slightly larger friction angle.