Together with the wall stiffness, a berm has the role of determining the stability of a temporary retaining wall during excavation. Especially in the case of a deeper excavation, the role of berm is very important.
In this study, the measurement data, obtained from the temporary retaining wall in the bermed excavation site in urban and the numerical analysis results, were used to investigate the effects of berm''s volume (width and slope), excavation depth and ground property on the maximum horizontal displacement of the temporary retaining wall.
The measurement data indicated that the berm was effectively restrained to the wall displacement. The wall displacement varied to the excavation depth and berm''s volume (width and slope). That is, as the excavation depth increased and the berm volume decreased, the wall displacement increased.
The finite element program (MIDAS GeoXD) was used to estimate the effect of berm on the displacement of the wall in detail. As a result of numerical analysis, it was found that the berm is effectively restrained to the wall displacement, which is the same result as the measurement data. The maximum wall displacement increased as the slope increased (steeper) and as the berm width decreased. In the case of the same berm condition, the wall displacement restrained as the ground property was better. As the excavation depth increased, to get the same effect of berm, the volume of berm needed to be increased.
A regression equation of wall displacement, with 93% of determination coefficient (R2=0.938), was constructed using the measurement data. An another regression equation with 70% of determination coefficient (R2=0.700) was also constructed using the numerical analysis results considering berm''s volume (width and slope), soil property and excavation depth.
A function of berm was evaluated using three methods; intersection point method, moment method, and friction angle method. The intersection point method took the virtual resistance location as the intersection between berm base and wall. This method overestimated the function of berm when the excavation depth increased. The moment method took the virtual resistance location as the first point of zero moment below the excavation base. It underestimated the function of berm when the excavation depth increased. The friction angle method took the virtual resistance location as the Lohemeyer''s method. Compared to other two methods, this method reasonably well-estimated the function of berm.
In addition, based on the results of intersection point method and friction angle method, new equations, which can estimate the berm width required to maintain the wall-stability during excavation, were proposed. These equations are so simple and can be used in practice easily.
A function of berm was also evaluated by comparison between the passive displacement of the berm and the maximum wall displacement. Both displacements were calculated using FEM program. If the passive displacement of the berm is larger than the wall displacement, the berm has no function of resistance of the wall.
In addition, to decide the berm function, a decision diagram was proposed as functions of berm width, berm slope, and excavation depth. This diagram was drawn based on the comprehensive analysis of numerical analysis data. The regime was divided into three regimes as two boundary lines, upper line with berm slope 1:05 and lower line with berm slope 1:1.0. The berm function presented good in the bottom regime, intermediate in the middle regime, and bad in the top regime.