Prediction of Seismic Lateral Earth Pressures on Broken-back Gravity Quay Walls Using Experimental, Analytical, and Numerical Approaches

• Seismic earth pressures on broken-back walls are assessed by pseudo-static, and dynamic analyses. • Numerical and analytical predictions are validated against 1g shaking table test results. • At maximum thrust, dynamic analysis yields more reliable predictions than pseudo-static method. • At post-seismic conditions, the pseudo-static method provides more precise predictions. • Applying correction factors improves pseudo-static predictions by an average of 3.3–6.7. Seismic lateral earth pressure is a key parameter governing the performance of retaining structures under earthquake-induced loading and serves as the basis for seismic analysis in relevant design codes and guidelines. Broken-back gravity quay walls, characterized by their unique rear-face geometries, exhibit distinct lateral pressure distributions resulting from the complex response of the backfill soil. In this study, the seismic lateral earth pressures acting on such walls are predicted using both nonlinear dynamic effective stress-based analysis and the Mononobe-Okabe (M-O) pseudo-static method, and are subsequently validated against corresponding 1g shaking table test data from the literature. For this purpose, explicit Lagrangian finite difference models are developed in FLAC 2D software for dynamic simulations, while a comprehensive spreadsheet-based analytical program is formulated for pseudo-static calculations. Nine wall configurations with varying geometries, seabed soil properties, and seismic excitations are examined at three stages: pre-shaking, at the instant of maximum lateral soil force on the wall, and post-shaking. Compared to experimental results, nonlinear dynamic analyses yield more accurate predictions at the moment of peak lateral thrust, whereas the M-O method provides closer agreement under post-seismic conditions. Applying the proposed correction factors to the M-O method enhances its accuracy in estimating maximum thrusts by an average factor ranging from 3.3 to 6.7, thereby improving its applicability for the seismic design of broken-back quay walls.