Abstract:

In engineering practice simplified methods are essential to the seismic design of embedded earth retaining walls, as fully- dynamic numerical analyses are costly, time-consuming and require specific expertise. Recently developed pseudostatic methods provide earth stresses and internal forces, even in those cases in which the strength of the soil surrounding the structure is not entirely mobilised. Semiempirical correlations or Newmark sliding block method provide an estimate of earthquake-induced permanent displacements. However, the use of these methods is hindered by uncertainties in the evaluation of a few input parameters, affecting the reliability of the methods. This study uses 1D site response analyses and 2D fully-dynamic finite element analyses to show that simplified methods can provide a reasonable estimate of the maximum bending moment and permanent displacements for stiff cantilever walls embedded in uniform sand, providing that a few input parameters are evaluated through semiempirical correlations and a simple 1D site response analysis.

Key words: embedded retaining wall, seismic design, finite elements, pseudostatic methods, Newmark, permanent displacement