Probabilistic Settlement Prediction for Neighbouring Footings at Different Spacing Distances in Rotated Spatial Anisotropic Multi-Layered Soil

This study develops a probabilistic settlement prediction model for neighboring footings in multi-layered soil under a rotated anisotropic spatial variability. The Random Finite Element Method (RFEM) is applied to predict settlement behavior over two layers, a strong clay layer overlying a weak clay layer, combining spatial uncertainty and anisotropic properties. Anisotropic random fields are introduced as rotated with various orientation angles and fluctuation scales to simulate variable soil conditions. Using the Mohr-Coulomb failure criterion within a Monte Carlo simulation, the settlement response is evaluated, modelling friction angle (Õ) and stiffness (E) as positively cross-correlated random variables employing Cholesky decomposition techniques with log-normal distribution patterns. The results show significant variability in settlement behavior due to rotated anisotropic fields, with the coefficient of variation (COV) increasing along rotation angles while mean settlement remains stable. These findings highlight the importance of anisotropic properties and random field orientations in predicting settlement for spatially variable layered soils subjected todifferent spacing of neighboring footing.

Keywords: Neighboring footing, Settlement, RFEM, Rotated anisotropy random field.