 |
| Titel |
Modelling of a viscoplastic granular column collapse and comparison with experiments |
| VerfasserIn |
Nathan Martin, Ioan Ionescu, Anne Mangeney, Francois Bouchut, Olivier Roche, Maxime Farin |
| Konferenz |
EGU General Assembly 2015
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 17 (2015) |
| Datensatznummer |
250112498
|
| Publikation (Nr.) |
 EGU/EGU2015-12658.pdf |
|
|
|
|
|
| Zusammenfassung |
| Landslides and, more generally, large scale granular flows, represent a wide variety of
geophysical flows also including mud or debris flow and snow avalanches. In a continuum
mechanics context, the accurate simulation of these flows strongly depends on the modelling
of their rheology and their boundary conditions, namely the sliding law and processes of
erosion. In particular the description of the static and of the flowing states of granular media
is still an open issue.
We focus here on the quantitative reproduction of laboratory experiments using a
mechanical and numerical model of dry granular flows with the so-called μ(I) rheology
associated to a Drucker-Prager plasticity criterion and a shear rate and pressure dependent
viscosity η(/¥D/¥,p). A Coulomb type friction law is considered at the base of the
flow.
The modelling is achieved in a finite-element context using the software FreeFem++. The
simulations are bidimensionnal and well reproduce quantitatively both the dynamical and
final shapes of the deposit. The effects of the sidewalls of the experimental channel, neglected
in 2D simulations, are investigated by introducing an extra term in the equations varying with
the inverse of the width of the channel, providing an enhanced agreement with the
experiments.
The numerical results show that the flow is essentially located in a surface layer behind
the front, while the whole granular material is flowing near the front where basal sliding
occurs. The static/flowing interface changes as a function of space and time, in good
agreement with experimental observations.
The resulting dynamic viscosity varies from very small values near the free surface and
near the front to 1.5Pa.s within the quasi-static zone. The results show a rather small yet
computationnaly expensive difference between a constant viscosity model and a μ(I)
rheology in the case of a rigid bed. This has important implication for application to real
geophysical flows. The role of an erodible bed in conjunction with this rheology is also
investigated. |
|
|
|
|
|
|