On 30th June 2012, Prof Kenichi Soga of Cambridge University, United Kingdom gave a two-hour lecture to about 90 participants on “Landslide-Physical and numerical modelling of large ground movements” at the Malakoff Auditorium, Wisma IEM in Petaling Jaya
The event was organized by CESIG, IEM and supported by UTM, GETD, TUSTD, ICE, SEAGS and AGSSEA. Prof Soga began the lecture by showing the debris flow of Colorado in 1997 and the submarine landslide in Hawaii in 2003. He then proceeded to show how the debris flow and submarine landslide can be modelled using centrifuge and numerical methods. Landslides are known to cause loss of lives and properties. It is known that these landslides can move rapidly (such as debris flow) and travel at long distances. The traditional way of evaluating the risk of landslide is by slope stability analysis and it’s Factor of Safety. This method however does not assess the extent and speed of ground movements. Prof Soga, using advanced computational and physical modelling demonstrated how the initial state of soil influences the movement of landslides. The speaker firstly showed some animations of large-scale model tests of sand embankment that failed from seepage in two different manners. The sand levee that was initially wet or damp failed rather abruptly and more extensively when compared to the dry embankment where gradual failure occurred near its toe.
Submarine landslide is known to move at massive volume and can travel at distance of up to 400km with surface gradient of only 0.1 degree. Such landslides can cause significant damages to marine environment and facilities (e.g. seabed pipelines). The key question in understanding submarine landslide impact is how to model the mass velocity and travelling distance correctly. Prof Soga presented the latest research work at Cambridge, using mini-drum centrifuge to model the submarine landslide flows. Actual flow velocity in the field can be predicted and factored using suitable gravitational scaling laws. It should be noted that the submarine landslides are more complicated to model due to the fact that the flow is also affected by water entrainment, frontal shear and hydroplaning, basal shear, flow thickness, water viscosity and etc.
Advances in numerical methods to solve continuum problems were introduced in the second half of the lecture. While many audiences may be accustomed to the commercially available programs such as Finite Element Method (FEM) or Finite Difference Method (FDM) which are all mesh-based type techniques, conventional methods have limitations in simulating very large ground deformation. Particle or mesh-free methods, or in this case called Material Point Method (MPM) can be used to model large ground deformation and simulate the landslide flow. Complex MPM consolidation formulation was developed to couple the effect of multi-phase soil and water movements. The model can also include partially saturated ground conditions.
An intriguing note was made in examples of levee failure experiment, in that, the mode of failure not only influenced by the shear strength properties of the soil, but also sensitive to the angle of dilation. The MPM method remarkably able to replicate various types of landslide movements, that is, from gradual retrogressive sliding to a more catastrophic failure (larger mass movements). In the case of modelling submarine landslide, water entrainment that affects the flow viscosity was included in the constitutive model.
The speaker highlighted some examples of new opportunities of slope monitoring technique using distributed optical fibre strain sensing. The sensor is known as Brillouin Optical Time Domain Reflectometry (BOTDR) is said to be capable of measuring continuous strain profile (eg. every 10 cm) along a standard 10 km long telecommunication optical cable. Some field instrumentation conducted by Cambridge Geotechnical Research Group on slopes near coastal area of UK was presented. An interesting point to note is that the fibre optics can be configured into various types of sensors depending on its applications. For example, for slope monitoring purposes, it can be used to monitor lateral displacements such as inclinometer, measure axial strain in soil nails and detect subsurface shallow movement of an embankment. Field instrumentation using BOTDR surprisingly drew the most interest from the audience during the question and answer session. The lecture was of great benefit to the participants and thanks is given to the organisers and speaker.