Effect of cyclic loading on stiffness degradation of Lime cement mixed clays for railway embankment

Abstract

Railway embankments in Sweden are subject to significant settlements caused by repeated cyclic loading from train movements, resulting in alignment issues and substantial maintenance expenses. The leading cause of embankment degradation is the dynamic response of natural soft soils to prolonged train-induced loading, leading to evolving soil stiffness properties over time. Ground improvement techniques, such as lime cementation of clays, have shown promise in enhancing soil conditions. However, the dynamic stiffness properties of these improved soils still uncertain under repeated loading cycles. This research project investigated the dynamic stiffness properties of lime cement mixed clays (LCC) and conducted a comprehensive comparative study with properties of natural clays (NC) by using a developed Python tool a significant advancement in the field. Furthermore, IQR method was utilized as a pre-processing step to filter outliers, enhancing the analysis’s effectiveness and reliability. The research involved undrained cyclic triaxial testing of clay samples at Chalmers Geotechnical Lab, with cyclic data collection using GDSlab software. Advanced data pre-processing techniques were employed to refine the cyclic raw data and extract critical cycle stage information for each test. The tool’s automation facilitates accurate computation of stiffness parameters, such as damping ratio (D), secant stiffness modulus (E), and secant shear modulus (G), under varying cyclic loading (qcyc) and stress-strain conditions. The investigation also addresses the uncertainties inherent in laboratory measurements by calculating propagation errors to determine their impact on the calculations of the stiffness parameter. The cyclic response under undrained conditions exhibits complex, non-linear stress-strain behavior. The findings revealed a slight trend in the stiffness parameters (D), (E), and (G) for LCC samples with two local cycle amplitudes, accompanied by an unclear trend with four global cycle amplitudes, which reflect the quality of those measurements. In contrast, the NC samples, which underwent four global cycle amplitudes, exhibited a distinct trend in the parameters, indicating a clear degradation within the sample.