Sliding stability analysis of gravity retaining walls using the pseudo-dynamic method

Abstract

The sliding stability of gravity retaining walls for the dynamic active case is studied using the pseudo-dynamic method. After rectifying the wall inertia equations, the pseudo-static method of Richards and Elms is compared with the present pseudo-dynamic method in the active case. Comparison of the two methods for the same combination of input parameters indicates that the pseudo-static method and the pseudo-dynamic method predict the same value of combined dynamic factor (FW) for the low range of horizontal seismic acceleration coefficient kh. However, for higher values of kh, the pseudo-static method overestimates the value of FW compared with the pseudo-dynamic method. Using the pseudo-dynamic method, it is observed that the maximum values of the stability factors (the soil active thrust factor FT, the wall inertia factor FI and FW) do not occur simultaneously during the application of input motions. When FW is at its maximum value (which is the sliding stability criterion), other factors are not at their maximum values, and thus multiplying the maximum of the partial factors causes overestimation of FW in the pseudo-static method. Comparing the predictions of the pseudo-dynamic method with the experimental results confirms that use of the pseudo-static method for seismic design of retaining walls may be over-conservative, especially for high values of peak ground motions.