Improvements to Pullout Failure Estimation in MSE Walls

ABSTRACT: The current practice of estimating the internal stability of Mechanically Stabilized Earth (MSE) structures entails use of empirical or semi-empirical methods to estimate the maximum reinforcement loads, which in turn is used to assess three potential failure modes, namely: (a) tensile rupture, (b) pullout and (c) connection failure of the reinforcement. These computations are performed at each reinforcement level, and the failure of the entire wall is assumed if the demand in at least one reinforcement exceeds the capacity of that failure mode. This paper discusses the limitations of assuming that the weakest reinforcement governs the stability of the entire reinforced soil mass, particularly in relation to displacement controlled failure mechanisms such as pullout. Other limitations of the current practice in estimating the pullout resistance are also discussed. To overcome these limitation, the Soil Reinforcement Interaction method (Weerasekara et al. 2017) is introduced, which explicitly accounts for the interaction between the reinforcement and backfill using a non-empirical analytical model. The method allows to compute a factor of safety for the entire reinforced soil mass and not only at each reinforcement level similar to the traditional practice. To further demonstrate the limitations of the existing methods and capabilities of the proposed method, two full-scale instrumented walls reinforced with smooth steel strips and high-strength nylon strips are discussed.