The open source mass flow simulation model DFGFWF



The open source mass flow simulation model


The project

r.avaflow is a transnational research project supported by the German Research Foundation (DFG), which is the lead agency, and the Austrian Research Fund (FWF) within the international program D-A-CH (between Austria, Germany and Switzerland) in the period July 2014 - June 2017. Explore the main objectives and methodologies of the project and learn more about its logical framework and the desired outcomes.


Catastrophic granular and debris flows occur in many mountain areas all over the world. Snow avalanches, rock or rock-ice avalanches, debris flows, lahars and pyroclastic flows are only some examples. An adequate management of the risk related to these phenomena requires a detailed and reliable analysis of the mechanisms involved in such processes. Even though much work has been done on this subject, and a number of physically-based models with a varying degree of complexity do exist, some problems still remain unsolved:

  • Flow over arbitrary topography, the role of viscous pore fluid or two-phase nature of flow, and particle and/or fluid entrainment have not yet been accounted for in an appropriate way.
  • Until now, no successful attempts have been made to build readily available open source applications of these complex models, which would be essential to make them available to a broader group of users in universities and public services.

This project offers an effective, innovative and unified solution to these two problems. It is therefore concerned with rapid geophysical mass flows, including avalanches and real two-phase debris flows, from a known initiation zone through the flow path along natural mountain topography into the deposition zone. For a given amount of mass and its distribution in the initiation zone, we are interested in the motion and geometric deformation along the track down the arbitrary topography, including the processes of erosion and deposition of mass along the track and the ultimate distribution of the deposited mass. This will also include the effect of dynamically evolving pore fluid pressure and/or evolution of the solid and the fluid components. An equally important focus shall be put onto the development of a user-friendly and freely accessible application of the developed model. This application builds upon the geoinformation software GRASS GIS, which is available as an open source product under the GNU General Public License. The new software is evaluated using physical model tests and well-documented mass flow events. These tests cover a broad range of processes and process chains including debris flows, debris avalanches and avalanches of snow or rocks.

Organigram illustrating the work packages. The abbreviations indicate the names of the project team members responsible for each work package: SP = Shiva P. Pudasaini, MM = Martin Mergili, WF = Wolfgang Fellin, JF = Jan-Thomas Fischer, JH = Johannes Hübl, KK = Karl Kleemayr, JK = Julia Krenn, HK = Helmut Kulisch, AO = Alexander Ostermann, GQ = Gustavo Queiroz de Oliveira, CS = Christian Scheidl, GS = Gregor Staggl.

Key objectives and methodologies

Sophisticated concepts exist for modelling the motion of granular avalanches and debris flows. However, none of these concepts accounts for the complexity of the processes or process chains to a sufficient degree, meeting all of the following criteria:

  • use of suitable single- and advanced two-phase mass flow models with explicit role of the pore fluid;
  • availability of a routine for entrainment and deposition of solid and fluid components;
  • use of sufficiently realistic basal friction models;
  • use of a high-resolution numerical scheme for flow over natural topography;
  • distribution as a user-friendly, freely accessible, GIS-based open source simulation tool.

We aim at developing a new, well calibrated, physically constrained, robust and advanced computational tool for a real two-phase granular and debris flow simulation, embedded in the framework of GRASS GIS. Some of the key features of the model shall be outlined here:

I. The models:

Including other pioneering models, this project offers an opportunity to extend the general two-phase mass flow model of Pudasaini (2012) by incorporating the curvature of the mountain topography (Fischer et al., 2012), erosion/deposition mechanics, the pressure-dependent yield strength (Domnik et al., 2013), variable bed friction angles, and sophisticated constitutive models for the solid and fluid phases. The extended model will incorporate several new aspects: (i) the entrainment and deposition at the base and side walls for solid and fluid phases; (ii) rheological changes; (iii) introduction of the yield-strength; (iv) the water runoff; (v) an internal singular surface.

II. Physical model tests:

As well documented experimental data of a real two-phase debris flows, with a largely varying degree of physical parameters, are not yet available, we develop and conduct experimental investigations in a physical model test (Bechteler and Kulisch, 1994; Kulisch, 2002). We conduct experiments with dry, partly saturated and saturated initially released debris masses, including bank/bed erosion and entrainment (Rickenmann et al., 2003). The debris mixture consists of water, granular sand, gravel and grit. The measurement engineering consists of high-speed cameras, ultrasonic devices, pressure gauges and force plates. The model tests will yield unique sets of data which are applicable for detailed validation and calibration of the extended general two-phase model, and will include the distribution of particle concentration, viscosity, non-Newtonian viscous shear stress, velocity, height, internal and wall friction, mobility parameter, virtual mass coefficient etc. The extensive set of data will significantly improve the quality of the numerical predictions of granular/debris flows, mud flows, mud floods and water waves.

III. A unified high-resolution computing scheme:

We develop an efficient and robust discretization for the emerging extended two-phase model for arbitrary three-dimensional topographies, with the proper choice of high order methods and good geometric properties. A robust numerical method built on reliable error estimates uses adaptivity in time and space for integrating the constitutive model. To accurately account for erosion/deposition an adaptive basal surface will be modelled. We employ modern material models from soil mechanics for the granular response by developing an appropriate interface with respect to hypoplasticity. A second interface will be developed for employing various basal sliding laws. This will be tested along with a recent innovative description of an automatically evolving general pressure-dependent yield-strength and Coulomb-viscoplastic sliding law (Domnik et al., 2013).

IV. Computing with pressure dependent material friction:

We simulate the complex behavior of granular material when it is mixed with water and pore-water pressure evolves due to shearing. More advanced models should be used, accounting for possible cyclic deformations. We include hypoplasticity with intergranular strains and barodesy (Kolymbas, 2009, 2012; Fellin and Ostermann, 2013). These aspects are important, particularly during the mass collapse, the transition and the depositional processes.