Journal of Geotechnical and Geoenvironmental Engineering · 2004 · 94 citations · 41 references
The study proposes a method to adapt reliability‑based design charts for mechanically stabilized earth walls to account for model bias and uncertainty. Using Bishop's simplified method and Monte Carlo parametric simulations, the authors varied key design parameters, performed additional simulations over broad ranges, and applied the resulting charts to design an MSE wall alongside deterministic methods. The analyses showed that backfill friction angle, reinforcement tensile strength, unit weight, surcharge, spacing, and length strongly influence failure probability, leading to reliability‑based design charts whose predicted reinforcement strengths lie between those from deterministic approaches.
A parametric study was conducted using Monte Carlo simulation to assess how uncertainty in design parameters affects the probability of internal failure of mechanically stabilized earth (MSE) walls. Bishop's simplified method was used to conduct the internal stability analyses. The results of the analyses indicate that the mean and coefficient of variation of the backfill friction angle, mean and coefficient of variation of the tensile strength of reinforcement, mean unit weight of the backfill, mean surcharge, mean reinforcement vertical spacing, and mean reinforcement length have a significant effect on the probability of internal failure of MSE walls. Based on the results of the parametric study, a series of additional simulations were conducted where the significant parameters were varied over a broad range. The results of these simulations were used to develop a set of reliability-based design (RBD) charts for internal stability of MSE walls. A method to adapt these charts to address model bias and model uncertainty is also presented. A MSE wall was designed using the RBD method and two other deterministic design methods. The required tensile strength of the reinforcement obtained from the RBD method fell between the strengths determined from the deterministic methods.
41