Analysis of Multivariate Regression and Deep Excavation in Expansive Soil Using PLAXIS 2D

Abstract

Expansive soil is unreliable for geotechnical purposes due to excessive compression, dispersion, collapse, low shear strength, low bearing capacity and high swell potential. Multivariate regression modeling has been adopted in this chapter to establish relationships between soil parameters. The aim being better understanding of complex uncertainties and predicting results, for which linear, parabolic and exponent models were designed. Deep excavation was simulated in expansive soil above sand layer using Diaphragm wall, soil anchors and geogrids in PLAXIS 2D, and numerical analysis was performed to predict the soil behavior during excavation with the aim of better visualization of discretized elements, calculation of stress, deformation in soil and more stable computing process. The primary objective of this chapter is to establish relationships between soil parameters by mathematical modeling and analyze practical situation by simulating staged deep excavation in expansive soil through numerical modeling.