The current design methods for the railway stepped reinforced soil retaining walls are still immature

which fails to satisfy the requirement of railway slope reinforcement practice. For the design of reinforced soil retaining walls

accurate determination of the potential failure surface is the key issue. However

though current specifications clearly address the shape and location of the potential failure surface

they limit their use only in single walls with a height of no more than 10 m. This paper attempts to identify the characteristics of potential failure surfaces in stepped reinforced soil retaining walls under railway loads by the use of model test. The model was designed with a large scale of 1:4 and the railway load was stimulated by a specific cyclic surcharge load. The shape and location of the potential failure surface were determined by the comprehensive analysis of the slope surface lateral displacements

vertical settlements at the slope crust and strains of the geogrid at different investigation points. The result illustrated: the potential failure surface in the lower stage wall is similar to that given by the “0.3H” method but has a deeper location and a more gentle inclination of the bottom part

which indicate a larger potential unstable area; while the shape of the potential failure surface in the upper stage wall is generally similar to the Ranking’s active rupture

however it thrusts into the lower stage wall rather than thrusts out from the foot of the upper stage wall. The test results and conclusions drawn in this study can provide theoretical reference for the design of railway stepped reinforced soil retaining wall.