Modelling of physical phenomena at water-soil interface

In the construction of structures that are built into the sea bed knowledge about the stability of the surrounding soil is crucial. For being able to place foundations in the seabed, excavations are necessary, which involve under water slopes. This was the motivation for the on-going research project, in which the focus lies on the stability of slopes under the influence of waves, the tide or ground construction work. However, once the structures are established, new challenges arise, two of which will be handled in the new research proposal: due to the skeleton-like shape of the soil, a fluid flow can establish inside it. This can lead to internal erosion (suffusion), which causes stability issues. An important application in this context is the flow around a sheet pile which will be studied closely. The second matter concerns the stability of dams: in many cases a dam separates two areas with different water levels. Between the region of high and low see level, seepage flow can be observed, and the developing phreatic line is of great interest. In the proposed project this topic will be investigated through the example of a vertical dam. The main issue will be the depiction of the air-water surface within the soil, which will require an advanced modelling of the surface tension. A combination of experiments and numerical simulations will be used to first develop suitable models and then deduce general statements about the effect of fluid or fluid-air flow within soil structures and around non-porous objects, i.e. sheet piles. As the soil structure is, in comparison to the overall problem size, quite fine, a multi scale approach will be used. Thus, for the numerical modelling of the above described processes coupled CFD-DEM simulations are used.

Goal of the DACH Cooperation project with FWF project number I3728 between Prof. Grabe (Institute of Geotechnical Engineering and Construction Management at Hamburg University of Technology (TUHH)) and Dr. Christoph Goniva (DCS Computing GmbH) was the investigation of phenomena at the water-air interface within soils as well as the suffusion of fine particle content. Altogether, four application cases were considered: the suffusion of fine particles in a bi-disperse particle packing, the desaturation of a soil column, the flow around a sheet pile and the formation of the phreatic line within a dam. For the investigations a combination of experiments, conducted by the project partners at TUHH and numerical simulations were used. For the latter we applied the Open Source software CFDEMcoupling - which combines the Open Source Computational Fluid Dynamics (CFD) code OpenFOAM and the Open Source Discrete Element Method (DEM) solver LIGGGHTS. The soil was represented as a packing of particles, which enabled the consideration of effects on a very fine-grained level. In the first project phase we focused on developments for an improved applicability of the software to large-scale cases: on the one hand we introduced a new, efficient algorithm for the generation of neighbour lists in polydisperse packings, on the other hand we implemented an insertion algorithm that allows for a much faster generation of dense packings of spheres. The packing generation is a preliminary step for all soil simulation cases, and its improvement leads to a strong decrease of the computational costs. In the next phase we developed and validated capillary force models that are required for a correct representation of the forces at the air-water interface in the presence of a solid phase with unresolved coupled CFD-DEM. Then, setups for all application cases were generated and according simulations were conducted for varying boundary conditions and different materials. The desaturation of the soil column and the formation of the phreatic line within a dam could be validated against models from the literature. For the validation of the flow around a sheet pile we could use experimental results from the project partners. In all three cases we found a good agreement between the simulation results and the validation data. After a feasibility study the suffusion setup was handed over to the project partner, who adapted it and used it for a thorough drag-force model investigation. Both cooperating parties presented the results and outcomes of the research project at international conferences and to specialists from both the field of geotechnics and numerical simulation.