Earthquakes have long posed challenges to geotechnical engineers, as serious destruction or harm can be caused at the surface. Especially cyclic liquefaction, the fluid-like behavior of soil under cyclic loading, is a phenomenon that has been extensively studied yet still needs m
...
Earthquakes have long posed challenges to geotechnical engineers, as serious destruction or harm can be caused at the surface. Especially cyclic liquefaction, the fluid-like behavior of soil under cyclic loading, is a phenomenon that has been extensively studied yet still needs more research. Two criteria are widely used to assess the potential of soils to liquefy, based on either the cone resistance qc or the shear wave velocity Vs. In practice, using liquefaction potential assessment criteria based on qc and Vs tends to lead to different outcomes. Because of this, the question arises what exactly contributes to the difference between the two. It is unclear whether potential differences can be attributed solely to differences in testing methods or whether the tested material primarily plays a role. The main question this research tries to answer is therefore: Can potential differences between results from qc-based and Vs-based liquefaction potential criteria be traced back to the intrinsic properties of the granular material?
Based on existing literature, fundamental principles behind liquefaction potential assessment criteria are investigated. These fundamental principles trace back to the physical meaning behind the parameters qc and Vs. Subsequently, data is gathered on four different types of sand: a rounded quartz sand, a compressible, angular carbonate, and two types of angular volcanic sand. This data originates from seismic cone penetration tests (SCPT), an invasive test measuring both qc and Vs. Based on four criteria, by Youd et al. (2001), Boulanger and Idriss (2014) (qc-based), Andrus and Stokoe (2000) and Kayen et al. (2013) (Vs-based), the resistance of the different sand types against liquefaction is determined in the form of the cyclic resistance ratio (CRR). Factors including compaction and aging are included in the analysis. Differences between different sand types are then interpreted considering the intrinsic properties of the granular material, like density, mineralogy, particle size and shape, gathered from existing laboratory data. Also, existing correlations between qc and Vs as well as new proposed ones are analyzed.
The fundamental difference between qc and Vs is that they represent behavior on opposite sides of the strain spectrum. Where qc is a large-strain parameter, Vs describes the soil’s behavior at very small strains. The results of the data analysis show large differences in qc and Vs across all sands, with carbonate sand showing low cone resistances while having high shear wave velocities. Both parameters appear to correlate better in quartz sand and the type 2 volcanic sand. This translates well to the corresponding CRR. Compaction affects both parameters, but seems to have a bigger impact on the qc than on the Vs, as well as the corresponding CRR. Ageing was analyzed over a two-year period, yielding a larger relative increase in Vs over qc. Based on literature, it is suggested that the CRR based on Vs represents the resistance of a soil against initial deformation, therefore the triggering of liquefaction. On the other hand, the CRR based on qc is more a reflection of the soil’s resistance against large strains, in other words, the extent of the consequences of the liquefaction event. In crushable soils, the qc appears to yield low values as the particles around the cone are crushed.
When considering how dependent the CRR is on the method used, it can be concluded that, when there is a good correlation between cone resistance and shear wave velocity, the CRR is to a small extent dependent on the type of liquefaction assessment method. However, when this correlation weakens, the corresponding CRR tends to deviate accordingly. Furthermore, it seems that the different liquefaction potential assessment methods are material dependent, and mainly the mineralogy and particle shape play a role. qc is mainly affected by soil mineralogy, as in carbonate sands particle crushing plays a big role. This can partly be accounted for with a Carbonate Correction Factor. Vs is significantly affected by particle shape and mineralogy. Mineralogy determines the cementation potential in a sand. qc-based criteria tend to be more material dependent than Vs-based criteria. Other factors that influence a correlation between qc-based and Vs-based liquefaction potential criteria are the compaction status (increasing the qc-based CRR more than the Vs-based CRR) and aging (increasing the Vs-based CRR more than the qc-based CRR). It is concluded that the intrinsic properties of the soil’s grains can explain to some extent its behavior when subjected to small or large strains, therefore explaining its resistance against liquefaction.
Questions remain around the use of the laboratory data that was gathered in the study, as testing samples are never from exactly the same location as the in-situ tests. This raises the issue on how representative the intrinsic particle properties are relative to the SCPTs used. The choice to limit the study to only the cyclic resistance ratio CRR is also discussed, as a safety analysis requires also the cyclic stress ratio CSR to get to a factor of safety, and the methods to determine susceptibility to liquefaction triggering that are evaluated in this thesis also differ in the procedure to determine the CSR. This study can therefore not be used for quantifying safety on site, but serves merely as a comparison of different liquefaction potential assessment methods, relative to each other. Finally, the legitimacy of the time-dependent increase in Vs is questioned with respect to liquefaction during an earthquake event, as the microstructure that was built up over time can be destroyed by the large imposed strains.
