Micro-mechanics of failure planes
Ricarda Gatter | Marum
Almost 20 years ago, in July 1998, a tsunami struck Papua New Guinea and devastated three villages, causing the death of over 2200 people. In 1996, a landslide initiated offshore Finneidfjord, Norway, and removed a 250 m long section of the main north-south highway, due to the slide’s retrogressive behaviour. Although these events highlighted the importance of submarine landslide research, the potential hazard related to submarine landslides was already recognised in 1929. A large slide and the flow of sediment it generated broke telecommunication cables off Grand Banks. However, to enable the assessment and effective mitigation of this hazard, one needs to understand the underlying mechanisms first. What causes submarine landslides? What are the pre-conditioning factors and triggering mechanisms? How do submarine landslides behave during runout?
I focus on the factors that pre-condition slope failure. Many studies have shown that basal failure planes of submarine landslides coincide with mechanically weaker layers embedded within the slope stratigraphy. Pore pressure fluctuations along potential weak layers, e.g. embedded volcanic ashes or fossiliferous soils whose particles can break down under loading, can decrease the shear strength of the sediments and hence, undermine slope stability. Although many studies have assessed the influence of sediment composition on the physical properties and shear strength of sediments, the processes occurring prior and during the initial failure are still poorly understood, as these cannot be directly observed or monitored. In this project, I will conduct a number of 3D numerical shear experiments, utilising granular simulation techniques (e.g., PFC3D), to investigate the failure processes at the initial stages of submarine landslides.
In a first step, I will compile a comprehensive dataset of µ-CT (micro- Computed Tomography) measurements of a wide range of sediment cores from different submarine landslides. These include cores from the Finneidfjord Slide (Norway), the AFEN Slide (UK), the Twin Slides (Italy), the Cap Blanc Slide (NW Africa), and the Tuaheni Slide (New Zealand). From this dataset, I will gather compositional and textural information of the material near the basal failure planes of the slides. The resulting information will directly act as input data for the numerical models.
By means of the 3D numerical simulations, which mimic geotechnical shear tests on a micro-(grain)-scaled level, I will investigate the fabric evolution of different materials under shearing with time. Of particular interest will be (A) a simplified ash and (B) a diatom sediment, and the changes in their shearing behaviour due to variations in the ash and diatom contents. The grain-scaled level is necessary in order to discern the spatial and temporal evolution of shear zones. Where does strain accumulate and where do basal failure planes form (above, inside, or below the weak layer)? This information is necessary to gain a deeper understanding of the failure mechanism and hence, the slide evolution itself.