Özet:

This study explores the impact of including the vertical spatial variability in effective stress friction angle of clay on the probabilistic analysis of deep excavations. The proposed methodology is demonstrated and verified by conducting random finite element modeling (RFEM) of an instrumented deep excavation project situated in Ankara, Turkey. The excavation has a depth of 20 meters and is supported by six levels of pre-stressed ground anchors. To simulate the vertical spatial variability of effective stress friction angle in the clay, Monte Carlo simulation method and the random field theory are employed. The simulated parameters are then inserted into the finite element model via Python programming language to analyze the probabilistic distribution of lateral deflections and bending moments in the drilled shaft wall. The results obtained from the Monte Carlo simulations reveal that the incorporation and selected value of spatial variability significantly impacts the resulting lateral movements, bending moments, and the probability of failure of the system.