Landslide mudflow behaviour: physical and numerical modelling

Abstract

Abstract Landslides constitute one of the natural phenomena that cause the most economic losses and deaths worldwide. After failure occurs, landslides can trigger mudflows. Understanding how mud is transported is very important in infrastructure projects that coincide with hillside areas due to the high risk of this phenomena occurring due to the high slopes, which can imply great risks and produce disasters, generating considerable costs. In this work, the evaluation of a mudflow is presented, from the execution of a scale experiment in the laboratory and its validation from numerical models, considering the material behaviour as a Newtonian fluid and as a non-Newtonian fluid. The physical model was developed using a 3m x 0.5m x 0.7m rectangular channel with dimensions, with slope control. A mud mixture composed of a silty material with 60% of moisture was tested producing a mudflow. Experimental tests were carried out with slopes of 5% and 10%. The numerical models were implemented in ANSYS FLUENT software. At first stage, the numerical model was calibrated with the results of the physical model with a slope of 5% and it was validated with the results of the model for the slope of 10%. Results of the numerical models were compared with the experimental results, and they have shown that these have a great capacity to reproduce what is observed experimentally. In addition, when the material was considerate as a Newtonian fluid, a similar behaviour was found respect to a mudflow as a non-Newtonian fluid, not finding considerable differences in the final deposition length of the flow. The simulation of mudflow, especially multiphase, using CFD is usually a complex process since the boundary conditions and physical and rheological properties of the soil must be correctly defined, considering the contribution of the solid fraction in the behaviour of the numerical model. Nevertheless, despite all the simplifications that a modelling of this type entails, the results are promising to improve the understanding of the phenomenon studied, and its application in risk assessment methodologies for mass movements and their derived effects.