Numerical Investigation of Collapse Effects in the Unsaturated Zone of Loess Soil: A Case Study of Kalaleh Loess
Numerical Investigation of Collapse Effects in the Unsaturated Zone of Loess Soil: A Case Study of Kalaleh Loess
Amirmasoud Naderi1 Peyman Alizadeh2 Vahid Beheshti3
1) M.Sc. in Geotechnical Engineering, Faculty of Technical and Engineering, Shahrekord University, Shahrekord, Iran
2) M.Sc. Student in Earthquake Engineering, Department of Civil Engineering, Shahrekord Branch, Islamic Azad University, Shahrekord, Iran
3) M.Sc. Student in Geotechnical Engineering, Faculty of Technical and Engineering, Shahrekord University, Shahrekord, Iran
IOI Code
:
AAGF-ABCEKZ
Publication :
9th International Conference on Applied Researches in Science & Engineering (carseconf.com/9th)
Abstract :
The collapse phenomenon in unsaturated loess soils represents one of the most serious geotechnical hazards, frequently triggering sudden settlements and structural damage to foundations. This study presents a detailed numerical assessment of collapse behaviour in the Kalaleh loess deposits. First, the geotechnical properties of undisturbed samples were determined through Atterberg limits tests (liquid limit 21–33 %, plasticity index 22–24 %), specific gravity tests (2.67–2.73 g/cm³), and consolidated-undrained (CU) triaxial compression tests (cohesion 0.58–1.08 kg/cm², internal friction angle 16.5–25°). A 12 m thick loess layer beneath a 2 m wide strip footing under a uniform pressure of 15 kPa was then simulated in PLAXIS 2D using a drainage-controlled wetting approach. Eighteen parametric scenarios were analysed, covering initial degrees of saturation from 0 % to 70 %, initial void ratios of 0.8 and 1.4, and wetting durations of 3, 10 and 30 days. The results exhibited a wide range of collapse indices (1.3–13 %), surface settlements (up to 228 mm) and maximum volumetric plastic strains (up to 15.1 %). Sensitivity analysis identified the initial degree of saturation as the strongest mitigating factor—reducing the collapse index by up to 80 %—while the initial void ratio emerged as the most significant aggravating parameter, more than doubling the collapse index under comparable conditions. Changes in pore-water pressure (up to 79 kPa) were recognised as the primary governing mechanism controlling collapse. These findings emphasise that neglecting the unsaturated zone can lead to prediction errors of several hundred percent in foundation behaviour. The high-quality dataset generated in this research provides a solid basis for developing artificial neural network models capable of rapid collapse prediction across the loess regions of Golestan Province and offers valuable practical guidance for the safe design of infrastructure in the area.
Keywords : loess collapse kalaleh loess unsaturated soil zone numerical modelling plaxis 2d hydro mechanical coupling