Pile installation in submerged sandy slopes

Abstract

Loose, sandy slopes situated or deposited under water may be susceptible to liquefaction-induced failure. In contrast to 'regular' localised slope failures, flow slides are diffuse, large-scale and potentially disastrous. Earthquake-induced ground vibrations currently dominate the liquefaction engineering scene. In reality, however, many liquefaction failures may be attributed to non-seismic sources of cyclic shear loading, albeit often in less dramatic context. Pile or sheet pile installation form an example. The construction of a new sea lock in IJmuiden brings this issue to the forefront and initiates a series of pile installation tests, conducted in several sandy deposits. The results of these tests are analysed and reveal (1) some key differences with seismic sources of vibration; (2) the importance of the duration and frequency of driving in residual excess pore water pressure (EPP) generation; and (3) the magnitude of spatial and temporal scales on which vibrations and EPPs generally act. The observations from the tests also form the basis for the establishment of a cyclic liquefaction model which simulates vibratory pile driving-induced EPP development. The model is validated using data from the IJmuiden pile installation tests. The cyclic liquefaction model is combined with a constitutive modelling approach for flow, or static, liquefaction behaviour. This enables the creation of a comprehensive strength framework, which allows representative strength parameters, based on the onset of flow liquefaction, to be derived for use in slope stability analyses. This strength framework is rooted in critical state soil mechanics and the related state parameter. A slope stability analysis procedure is advocated in which priority is given to pre-pile installation liquefaction analysis. Given a satisfactorily stable slope, pile installation effects are incorporated, where EPP generation and migration of pore water are considered the dominant contributors to failure. The procedure is applied to a fictional reference slope, where results show significant reduction in safety against global failure during driving. Considering three-dimensional drainage effects, however, suggests that pile installation may only have a minor effect on slope stability, depending on the failure mechanism and volume under examination. From the example stability analysis, some recommendations are made with regard to prevention and mitigation of piling-induced slope failures. Finally, in terms of further study, it must be noted that several major simplifications and assumptions underlie the cyclic liquefaction model and the corresponding method for slope stability analysis, which may be improved upon with more rigorous constitutive modelling and numerical analysis.