Submarine slope instabilities are considered one of the major threats for offshore buried pipelines. This paper presents a novel method to evaluate the ultimate pressure acting on a buried pipeline during the liquefaction of an inclined seabed. Small-scale model tests with pipes buried at three different embedment ratios have been conducted at an enhanced centrifugal acceleration condition. A high-speed, high-resolution imaging system was developed to quantify the soil displacement field of the soil body and to visualize the development of the liquefied zone. The measured lateral pressures were compared with the hybrid approach proposed for the landslide–pipeline interaction in clay-rich material by Randolph and White (2012) and Sahdi et al. (2014). The hybrid approach is proved to be able to predict later pressures induced by the movement of (partially) liquefied sand on buried pipelines. It is found that the fluid inertia (fluid dynamics) component plays an important role when the non-Newtonian Reynolds number >~2 or the shear strain rate > 4.5 × 10⁻² sec⁻¹.

The scaling laws for the centrifuge modelling of the initiation and propagation of static liquefaction in submerged slopes are investigated in this paper. A theoretical model is developed to analytically determine the scaling factor of fluid viscosity in simulating the onset of static liquefaction by detailed analysis of the hydromechanical processes at the grain scale. Based on this, a fluid with a viscosity of N-times that of water (N-fluid) is suggested, where N is the geometrical scaling factor in centrifuge modelling. A fluid with a viscosity of N-times that of water (N-fluid) was adopted for simulating dynamic events in the centrifuge; N-fluid is used and suggested by previous researchers. Centrifuge tests were designed to examine and verify the scaling factors for pore fluid viscosity in simulating the onset of static liquefaction and the post-liquefaction behaviour of subaqueous landslides. These tests were performed at 10g, 30g, and 50g conditions, with N-fluid, N-fluid or water, where g is the Earth’s gravitational acceleration. Results confirm that the correct scaling factors (prototype/model) for pore fluid viscosity are 1/N and 1/N for investigating the onset of static liquefaction and the post-failure behaviour of the submarine slopes, respectively.

The assessment of the potentially destructive impacts of subaqueous landslides on offshore pipelines is required when the pipeline route passes through zones with a risk of mass movements. Therefore, quantifying and evaluating the ultimate load/pressure acting on the pipeline is one of the key factors in geotechnical safety design of the pipeline. One of the triggers of subaqueous soil mass movements is the monotonic loads, which induce the trigger relative displacement between a soil layer and a pipe under both drained and (partially) undrained conditions. Two approaches based on geotechnical and fluid dynamics perspectives have been proposed for estimating the ultimate load/pressure for different stages of a submarine landslide. Traditionally, the former method focuses on the analysis of pipelines installed under flat seabed experiencing relative movements to the surrounding soil, whereas, the latter method focuses on the behaviour of pipelines laid on the surface of the seabed and subjected to debris flows. However, offshore pipelines are often buried under the seabed, which is not always flat and has a modest inclination in some cases. This engineering condition normally differs from that of the simplifying assumptions and boundary conditions (such as seabed inclination, and soil strength) commonly imposed to the geotechnical and fluid dynamics approaches. Accordingly, a better understanding of the soilpipeline interaction when the pipelines are buried in subaqueous slopes is essential for evaluating the ultimate load/pressure that would be caused by the slope failures. This thesis presents a research effort on investigating the soilpipeline interaction during subaqueous slope failures using advanced physical modelling. In this research, the experiments can be divided into two main groups according to the soil drainage conditions. The first group of tests were carried out in the drained condition by using dry sand as the soil material for the slopes. The pipe was buried at 5 different locations inside the slopes to study the pipe burial position and pipe embedment ratio effects on the ultimate pressure during slope instability. Particle image velocimetry analysis was conducted to study the pipe movement and slope failure mechanisms. The results of these tests reveal that the slope angle and the pipe distance to slope crest play significant roles on the ultimate loads acting on the pipe.

The undrained shear strength of naturally deposited, consolidated and saturated clay usually increases with depth due to the overburden pressure and the decreasing water content. Nonetheless, calculations of the bearing capacities of foundations under undrained conditions are usually done using the approach of Prandtl, which assumes constant values of the undrained shear strength. Analytical models using the kinematic element method show that the failure mechanism and the bearing capacity are both influenced by the increasing undrained shear strength; therefore, Prandtl's method might overestimate the bearing capacity of a shallow foundation assuming an average value of the undrained shear strength. Centrifuge tests of shallow strip footings on kaolin clay were conducted to validate the analytical model in terms of bearing capacity and failure mechanism. The estimated bearing capacity of the shallow foundation based on analytical models show good agreement with the centrifuge test results. However, the shape of the failure mechanisms is considerably different. In the analytical models, the development of the failure mechanism is estimated based on the assumption of a perfectly plastic behaviour, developing a rigid-body collapse. This behaviour was not observed in the physical models due to the development of the failure mechanism.

Sand erosion and scouring caused by waves and marine currents result in gradual increase of local seabed inclination and formation of slopes around hydraulic structures and offshore foundations. During this process, shear stresses in the soil body increase monotonically which may lead to static liquefaction and damage of the adjacent offshore infrastructure. This paper presents the details of a newly developed static liquefaction triggering actuator to be used at an enhanced gravity condition in a geotechnical centrifuge. This actuator simulates the steeping process of submarine sand layers due to scouring and enables the investigation of failure mechanisms in submerged slopes. The details of the centrifuge test set-up designed and constructed to simulate the process of triggering static liquefaction in loose sand layers are presented. Furthermore, the performance of the novel integrated model preparation facility using sand fluidization is explained. The set-up was used to conduct several centrifuge tests at four different slope steepening rates to investigate the slope steepening rate effects. Moreover, the effect of viscosity of the submerging pore fluid on the behaviour of the slopes at the onset of failure is investigated. The Coriolis effect on loose saturated sand samples during increase of g-level is examined as well. Results show that the built-up of pore pressure due to local shear deformations can be detected and considered as one of the triggering mechanisms of this kind of submarine slope instabilities.

Significant forces can be applied to embedded pipelines in sloping grounds due to soil instabilities, which potentially might lead to leakage of hazardous fluids into the environment. The soil-pipeline interaction in sandy slopes has been investigated experimentally using small-scale physical models tested in geotechnical centrifuge. A novel method is developed in this paper to estimate the ultimate external forces, induced by slope failures, acting on buried pipes at various locations inside the slope. Instabilities were triggered by surcharge loading on the slope crest in the centrifuge tests. Six dense coarse sandy slopes were tested with different pipe locations with respect to the slope crest. Moreover, two medium dense fine sand slopes were tested in the same manner to study the effect of the grain size distribution on the soil-pipe interaction. The external forces on the pipe induced by the surrounding soil movements were calculated based on the measurements of four strain gauges installed on the pipe. The shape of failure surface and pipe movements were monitored with the aid of advanced image analysis techniques. The results indicate that a buried pipeline has the potential to affect the slope failure mechanism. Normalised force-pipe displacement relationships were derived and compared to the estimation methods suggested in previous studies, which were mainly done on pipes installed in flat grounds. A new prediction method is introduced in this study, which considers the pipe burial distance to the slope crest. Moreover, the slope angle effect on the ultimate force applied to the pipe is also investigated, and a generalised formula is developed. Finally, two examples of the application of the new method are presented for pipelines installed at the toe of two large-scale subarial and submarine slopes.

The profile of the undrained shear strength of kaolin clay samples has been determined using an automated shear vane device. The automated device can monitor and register the changes in the mobilised shear resistance as a function of shear strain, while merely peak undrained shear strengths of the sample can be determined using conventional vane shear tests. The clay samples were prepared using centrifuge consolidation of slurries at 30 g and 90 g. The results on the impact of different su-profiles to the load displacement curves and bearing capacities of shallow foundations which were tested at 30 g on normally consolidated and overconsolidated kaolin samples are shown. The reliability of field-vane-shear tests on borehole samples is also discussed.