Performance Analysis of Axially Loaded Secant Pile Wall Embedded in Sand: An Experimental Investigation

Abstract

Abstract In urban environments, temporary excavation support systems (ESSs) are intensively recommended during the construction process of structures with underground levels to preserve nearby structures and maintain the excavation sides. Once the foundations and basements are constructed, these systems are rendered useless. As a result, integrating the temporary ESS into the building foundation may have significant benefits. Therefore, the main aim of this paper was to investigate the behavior of Secant Pile Walls (SPWs) through fifteen model tests with an acceptable scale on an axially loaded SPW embedded in medium and dense sand. This study considered several factors to define wall behavior, such as normalized lateral deflection ( δ h / H t %), the vertical deflection of the SPW ( δ vw / H t %), vertical ground settlement ( δ v / H t %), and settlement influence zone ( D o ). These factors were investigated and analyzed under the influence of a set of parameters including normalized penetration depth ( H e / H c ), sand relative density ( D r ), and surcharge load density ( W sur ). The findings demonstrated that SPWs had structural and overall stability features to withstand lateral earth pressures as well as applied axial loads. Generally, increasing the H e / H c ratio further than a limit value of 2.0 for the same surcharge load had a limited impact on the ultimate axial capacity, particularly in the case of dense sand. The location of the pivot point ( ε ′) extended from 0.24 to 0.41 H e from the wall tip, with a mean value of 0.34 H e and 0.29 H e for the values of D r = 80 ± 2%, and 60 ± 2%, respectively. Other issues were also discussed for selected samples, including an analysis of the wall's bending moments and any potential wall buckling. Finally, to correlate the experimental data with the theoretical values, a modification factor for the pile static formula was developed by using nonlinear regression analysis with a significant prediction accuracy with an R 2 of 0.94.