Granular Flow Experiments for Validation of Numerical Flow Models

By Amy Webb, Marcus I Bursik1

1. University at Buffalo, SUNY (UB)

Published on


Many types of geophysical flows such as debris avalanches and pyroclastic flows pose significant danger to both people and property. Understanding granular avalanche behavior and particle interaction during the course of a flow is very important in the development of granular flow models. Since it is not always convenient or possible to use actual events to validate flow models, model results are compared to laboratory experiments. Laboratory experiments permit control over parameters such as material properties and bed geometries, which allow for easy comparison of flow theory with experiments.

The research involved the development of granular flow experiments to demonstrate the strengths and limitations of the theoretical flow model Titan2D. In the first experiment, avalanches were initiated down a smooth, inclined plane. The resulting flow propagation and deposition were compared to Titan2D simulations. The Titan2D model performs well in replicating the geometric propagation, position and timing of dry granular avalanches down smooth inclined beds.

In the second experiment, avalanches were initiated down a rough inclined plane. The rough plane was covered with an erodible bed and the resulting mobilization and deposition of static particles was measured and analyzed to better understand erosion rate over time. The data gathered from these experiments will be used to test the hypothesis that erosion can be incorporated into the Titan2D model by formulating criteria similar to Shield’s stress criteria used in fluidized flow. 

In the last experiment, photoelastic discs were used in two-dimensional realizations of static non-cohesive granular systems. Measurements were taken of the resulting filamentary stress/force networks within the discs to gain a better understanding of the force distribution occurring within dynamic granular systems as well as to validate the use of depth-averaging flow variables inherent in the Titan2D model. By analyzing the orientation of stress chains that developed in the photoelastic discs, it was observed that for the range of angles tested, depth-averaging flow variables is a viable option.

This contribution includes an explanatory manuscript, and image and text data for the first and second experimental sets. 

Cite this work

Researchers should cite this work as follows:

  • Amy Webb; Marcus I Bursik (2016), "Granular Flow Experiments for Validation of Numerical Flow Models,"

    BibTex | EndNote