Stability analysis for road-cutting : Review, recommendations and examples

The construction works of the railway connection Seinäjoki-Oulu in Pohjanmaa revealed several stability problems and failures in the road cuttings of underpasses particularly in soft clays. During design stage, cutting slopes have been checked against stability using the slices method (LEM) and they have then fulfilled the stability requirements. This report reviews the fundamental principles of limit equilibrium and choice of shear strength parameters as employed in the slices method for the evaluation of road-cutting stability. The provided data and design studies show that, generally speaking, the designer has followed the common practical rules in checking the stability of the designed soil cutting slopes. This study aims at identifying the possible reasons (uncertainties) behind the overestimated safety and to conclude with design recommendations to reduce the possibilities of future failures. After recapitulation on Mohr-Coulomb failure criterion and the effect of the followed stress path on the soil strength at failure, the report discusses the choice of the shear strength parameters in undrained and drained conditions. This is followed by an overview on the method of slices, its different versions and how to incorporate groundwater flow in the analysis. Based on the findings of the review, two case histories (in Tuuliharju and Zatelliiti locations) that suffered failure in reality are re-evaluated for stability. The evaluation shows that the inappropriate design is related to the inaccurate choice of shear strength parameters and also to the inadequate modelling of pore water pressure during short and long-terms. The report concludes with recommendations to be followed during the design phase. The recommendations cover the choice of suitable strength parameters for undrained and drained conditions. They also give guidance on the determination of the length of undrained phase, groundwater flow calculations and suitable tests for shear strength estimation. One of the most important recommendations in case of cutting is that the employed shear strength parameters should be estimated using the direct simple shear test or alternatively based on the reduced values of the in-situ vane shear test. The shear parameters derived from the triaxial compression test tend to be non-conservatives whereas those from the triaxial extension test are conservative. The simple shear test (or reduced vane shear) gives average shear strength that is appropriate in the cutting situation. On top of that, the reasonable estimation of soil permeability and the transient changes in pore water pressure during and after cutting shows to be extremely important, especially for drained analysis. For example, the calculations show that in the case of clay with low permeability (K<1.0x10-9m/s), the commonly adopted dewatering system in practice is not efficient and the water level stays almost at its initial conditions after two months of pumping. Example calculations for cutting in three different soil-layering profiles are documented. These examples are sought to be representative of practical cases and illustrate how to implement the recommendations given in this report to improve the safety evaluation of the cutting during design phase.