| Geosynthetic-reinforced slopes are conventionally designed using methods based on limit equilibrium. Use of these methods requires the distribution of the reinforcement tensile forces with height as input information to estimate the factor of safety against internal stability. A linear distribution of reinforcement tension with height, with zero tension at the crest and maximum peak tension at toe elevation, has often been assumed. Although this assumption may be appropriate for the design of vertical geosynthetic-reinforced walls, little evidence has been collected justifying this distribution for the design of geosynthetic-reinforced slopes. In addition, current methods use the ultimate tensile strength, rather than mobilized reinforcement tensions, as a basis for selection of the tensile forces used as input in the analyses. A combination of centrifuge testing and digital image analysis is undertaken in order to obtain the strain distribution within geosynthetic-reinforced slopes under pre-failure conditions. Specifically, digital image analysis techniques are used to determine the displacement distribution along reinforcement layers in reduced-scale models subjected to increasing g-levels. A sigmoid function was useful to fit raw displacement data and estimate the strain distribution along reinforcement layers. Analysis of the reinforcement strain results shows that the distribution of peak tensile forces with height is proportional to the overburden pressure defined by the vertical distance below the slope, which leads to a maximum peak tension that is not located at toe elevation. Instead, the location of the maximum peak reinforcement strain was found to be a function of the slope inclination. The pattern of reinforcement peak strain with height obtained for pre-failure conditions was found to be similar to that obtained for failure conditions. While, factors of safety calculated assuming a constant reinforcement tension distribution can be used as a design stability index, use of the actual mobilized tensions in limit equilibrium analysis was found to lead to factors of safety of approximately one for load conditions well below failure. |
