Deepwater offshore oil and gas developments require an assessment to be made of the risk of infrastructure damage from submarine slides. The likelihood and magnitude of submarine slides, and the consequent impact loading on seabed infrastructure in the path of the debris from the slide, must be estimated. Export pipelines are especially vulnerable to impact from submarine slides, because of their length and the need to cross canyons and other seabed features that are potential paths for the flowing debris. Characterising the debris material represents a particular challenge, as the original soil, which is typically characterised using conventional geotechnical methods, evolves through remoulding and water entrainment into a viscous fluid. Because of this transition from soil to fluid, characterisation of the strength of flowing fine-grained sediment has been addressed separately within a soil mechanics framework and a fluid mechanics framework, resulting in two different approaches for expressing the strain-rate-dependent strength of debris flows, and the consequential impact loads on pipelines. In this paper we compare the two approaches, and show that the geotechnical characterisation of fine-grained sediments can be extended into the liquid range in a continuous fashion. This is supported by a series of undrained shear strength measurements on two different remoulded soils, from fall cone tests, vane shear (including viscometer) tests, T-bar and ball penetrometer tests. Analysis of the results shows that the variation in shear strength over the solid and liquid ranges can be described by a unique function of water content, for a given soil. Furthermore, the effects of rate of shearing are well captured by a dimensionless function of the normalised strain rate. The geotechnical approach also accounts for the observed strength reduction due to intense shearing.

SUMMARYEffective capabilities of combined chemo-elasto-plastic and unsaturated soil models to simulate chemohydro-mechanical (CHM) behaviour of clays are examined in numerical simulations through selected boundary value problems. The objective is to investigate the feasibility of approaching such complex material behaviour numerically by combining two existing models. The chemo-mechanical effects are described using the concept of chemical softening consisting of reduction of the pre-consolidation pressure proposed originally by Hueckel (Can. Geotech. J. The equilibrium equations combined with the CHM constitutive relations, and the governing equations for flow of fluids and contaminant transport, are solved numerically using finite element. The emphasis is laid on understanding the role that the individual chemical effects such as chemo-elastic swelling, or chemo-plastic consolidation, or finally, chemical loss of cohesion have in the overall response of the soil mass. The numerical problems analysed concern the chemical effects in response to wetting of a clay specimen with an organic liquid in rigid wall consolidometer, during biaxial loading up to failure, and in response to fresh water influx during tunnel excavation in swelling clay.

The clay-interface shear resistance is an important parameter for the design of offshore pipelines, which slide on the seabed as a result of thermally-induced expansion, contraction and lateral buckling. This paper presents a methodology for characterising the clay-interface resistance and quantifying the effect of drainage and consolidation during or in between shearing episodes. Models for describing the clay-interface resistance during planar shearing are presented and compared to test data for a range of drainage conditions from drained to undrained and including the case of episodic consolidation. The test data are from two series of interface shear box (ISB) tests carried out on marine clays. The effects of normal stress level (in the low stress range), over-consolidation and interface roughness are also examined. 1 INTRODUCTION Soil-interface resistance is a relevant design parameter for a range of geotechnical structures, e.g., piles, retaining walls, submarine pipelines, skirted foundations, offshore gravity structures and mat foundations. Studies of soil-interface behaviour have been conducted for a variety of soil types (e.g., Potyondy 1961, Yoshimi & Kishida 1981, Lemos & Vaughan 2000). These studies are based on experimental investigations using interface shear box (ISB) tests or interface ring shear tests, which have historically been performed as drained tests. Databases of results have been assembled by Ramsey et al. (1998), Jardine & Chow (2005) and Eid et al. (2014) (with the latter focusing on low stress levels) with the aim of identifying correlations between soil index properties and soil-interface shearing characteristics. Some studies have considered tests conducted at different rates of shearing, corresponding to various drainage conditions (e.g, Lehane and Jardine 1992, Tika et al. 1996, Ganesan et al.

SUMMARYA solution to the problem of freezing of a poroelastic material is derived and analysed in the case of one-dimensional deformation. The solution is sought within the framework of thermo-poroelasticity, with specific account of the behaviour of freezing materials. The governing equations of the problem can be combined into a pair of coupled partial differential equations for the temperature and the fluid pressure, with particular forms in the freezing and the unfrozen regions. In the freezing region, the equations are highly non-linear, partly due to the dependence of thermal and hydraulic properties on water saturation, which varies with temperature. Consequently, the solution is obtained through numerical methods, with special attention to the propagation of the freezing front boundary. The response to one-dimensional freezing is illustrated for the case of cement paste. Finally, the influence on the solution of varying selected parameters is analysed, such as the temperature boundary conditions, the parameters characterizing the geometry of the porous system, the ratio of fluid and thermal diffusivities, and the rate of cooling applied at the freezing end.