On the role of intrinsic soil properties in flow liquefaction susceptibility

Abstract

Soil liquefaction is a phenomenon in which an otherwise stiff, loosely packed, cohesionless soil loses its strength and behaves behaves like a viscous liquid in response to a change in stress conditions. This study focuses on liquefaction under monotonic load (i.e. static liquefaction) and not cyclic liquefaction. While the role of state variables such as relative density in liquefaction is well established, the importance of intrinsic soil properties (ISPs) is less clear. ISPs include grain size gradation, mineralogy and grain shape The critical state soil mechanics framework can be used to link these properties to liquefaction susceptibility. One approach to do so is the "Relative Contractiveness" (RC) concept proposed by Verdugo & Ishihara (1996). This thesis investigates the role of ISPs in soil liquefaction and tests the RC concept through a combination of statistical analyses, four case studies and a new experimental study.

The statistical analysis shows that an increasing fines content generally leads to greater relative contractiveness and especially at lower stress levels, indicating increased sensitivity to liquefaction. Particle shape plays a multi-faceted role in liquefaction susceptibility, as increased angularity may increase compressibility but also increase resistance to particle rotation and hence reduce the likelihood of flow behaviour. The mineralogy of soils was difficult to statistically analyse as the information is usually not given, but extra care should be taken when dealing with sands that are not made of quartz, as most index methods are based on quartz.

The case studies exemplified varied applicability and benefit per case. The Ijmuiden case demonstrated the limitations of field tests and critical state determination. It did indicate medium to high relative contractiveness for the tested soils. The Nerlerk berm failure demonstrated the importance of fines content in liquefaction susceptibility, as only the finer of the two soils used for the hydraulic fill liquefied. However, the geometry and differences in deposition method also played a role. For the Hollandsch Diep case environmental factors are ought to play a more important role in liquefaction rather than that the soil is intrinsically exceptionally susceptible to liquefaction. The Bangabandhu bridge case highlighted the limitations of compressive loading based methods as the soil was particularly weak in tensile loading. It also highlighted the importance of mineralogy and grain shape, as the presence of plate-like micaceous particles drastically reduced its strength.

The new experimental study investigated a soil from the Eastern Scheldt estuary in the Netherlands, a region historically notorious for liquefaction flow slides. Surprisingly, the sampled soil was not prone to liquefaction at all, showing strong dilative tendencies under triaxial compression.

In conclusion, this study suggests that the relative contractiveness concept could be used as a screening method for assessing liquefaction risk, rather than a deterministic method for designing parameters. However, further studies with extensive and consistent material characterization and critical state determination are needed to verify the validity of the relative contractiveness concept. Discrete element modelling of soils could also provide future opportunities for advancing our comprehension of the role of ISPs in liquefaction susceptibility.