PARTICLES 2023

2D combined DEM-LBM modelling of submerged sinkhole occurrence during a massive flood event

  • Philippe, Pierre (INRAE - Aix Marseille University)
  • Fan, Jianhua (Jilin University)
  • Luu, Li-Hua (INRAE)
  • Noury, Gildas (BRGM)

Please login to view abstract download link

Surface water inflow is a factor accelerating the formation of karst sinkholes. An exceptionally heavy supply of water to the ground surface can indeed lead to a fully saturated situation and subsequent gravitational drainage that can locally affect soil stability. For instance, this is most likely what happened during the massive flood of spring 2016 in the Orléans region (France), resulting in an extremely high occurrence of sinkholes. In the very limited literature on this subject, it is considered that there are two types of sinkhole formation processes: dropout sinkhole due to an upward cavity expansion from the underground, with successive partial collapses of the roof; subsidence sinkhole when the whole soil layer is destabilized and collapses through the karst conduits. We present here a simplified micro-mechanical study of such processes based on numerical simulations combining both the Lattice Boltzmann Method (LBM) and the Discrete Element Method (DEM) in a 2D modelling of submerged sinkhole collapse including a specific model for bond cohesion between soil particles. The study allowed phase diagrams to be drawn and the expected scenarios of dropout and subsidence sinkholes to be placed in them, as well as a third situation corresponding to the formation of a stable cavity. Then, we focus specifically on the case of subsidence sinkholes, which actually correspond to the emptying of grains through an orifice but in submerged condition and with inter-particular cohesion. For illustration, two comprehensive analyses about the influence of the fluid/grain coupling and of the cohesive force between grains will be presented.