2009 Portland GSA Annual Meeting (18-21 October 2009)

Paper No. 6
Presentation Time: 9:45 AM

A LOAD MODEL FOR THE DESIGN OF FLEXIBLE DEBRIS FLOW BARRIERS


WENDELER, Corinna, Geobrugg AG, Aachstrasse 11, Romanshorn, 8590, Switzerland, BARTELT, Perry, Institute for Snow and Avalanche Research SLF, Flüelastr. 11, Davos, 7260, Switzerland and SHEVLIN, Tim, Geobrugg North America, LLC, Salem, OR 97302, OR 97302, corinna.wendeler@geobrugg.com

Instrumented flexible debris flow barriers have been installed at the Illgraben debris flow research observation station, located in Canton Valais, Switzerland. Analysis of cable forces, coupled with debris flow pressure and velocity measurements, during several real scale debris flow events have allowed us to construct a load model for barrier design. The field tests, with additional laboratory experiments, led to the development of a multi-stage surge model. We found that during the stopping of the first surge, the cable forces can be effectively modeled assuming a dynamic impact pressure working in combination with the hydrostatic loading. However, the complex surge dynamics of the fill process can lead to higher cable forces and possible barrier overflowing. In this case the associated dynamic load depends on the velocity and mass of the first surge. Its drag coefficients, depending on the debris flow material characteristics, were determined. During debris flow events, the total pressure distribution on the ring-net can be approximated by iteratively tracking the second and following surges over the original deposits. If the barrier is completely filled, the next surge is overflowing the filled barrier and is influencing the net with its normal and shear force component. The hydrostatic pressure and the additional weight of the overflowing surge are time dependently reduced by compaction and drainage.

Both field and laboratory test indicate a non-constant dry Coulomb friction coefficient. Since we measured both shear S and normal force N at the base of the debris flow during the experiments, we could calculate effective friction coefficients S/N. We found higher friction values at the front in comparison to the tail of the flow.