HR: 0830h
AN: H31C-0478    [PDF]
TI: Erosion and its Effects on Geophysical Granular Avalanches
AU: * Webb, A
EM: starphish@yahoo.com
AF: SUNY at Buffalo, Dept. of Geology 876 Natural Sciences Complex, Buffalo, NY 14260 United States
AU: Bursik, M
EM: mib@buffalo.edu
AF: SUNY at Buffalo, Dept. of Geology 876 Natural Sciences Complex, Buffalo, NY 14260 United States
AU: Patra, A
EM: abani@eng.buffalo.edu
AF: SUNY at Buffalo, Mechanical and Aerospace Engineering 318 Jarvis Hall, Buffalo, NY 14260 United States
AU: Nichita, C
EM: cnichita@acsu.buffalo.edu
AF: SUNY at Buffalo, Mechanical and Aerospace Engineering 318 Jarvis Hall, Buffalo, NY 14260 United States
AU: Pitman, B
EM: pitman@buffalo.edu
AF: SUNY at Buffalo, Dept. of Mathematics 244 Mathematics Building, Buffalo, NY 14260 United States
AB: Many types of extreme geophysical flows such as debris avalanches and pyroclastic flows can be classified as granular avalanches. While sedimentation from these flows and their resulting deposits have been the subject of intense study for the past half-century, the effects of erosion have been difficult to ascertain. Despite the known relationship between slope angle and thickness of a stable layer of loose debris on a slope, $h_{stop}$, which requires erosion during flow events, the positive change in volume of geophysical granular flows during transport has not been systematically studied. We present the results of experiments on the propagation of erosive granular flows. We use the experimental results to test a model that simulates granular avalanches in which erosion and deposition cause significant changes in mass and momentum during flow. Granular flow experiments were conducted under a variety of initial conditions to gain a quantitative understanding of erosion and its effects on propagation. Sand flows were released on a masonite plane to which 36 grit sandpaper had been glued. The masonite plane measured 95 cm times 60 cm. The plane was tilted at angles of 29.4 -- 36.5 degrees. Particles of sand grains sieved to 2 and 2.5 phi were used. The 2 phi particles were dyed blue to aid in visualization. The angles of repose for the 2 and 2.5 phi particles were measured in two types of slumping experiments yielding angles of 34 degrees for the 2 phi particles and 35 degrees for the 2.5 phi particles. The 2.5 phi particles were poured onto the plane before each experiment to provide an erodible surface. The thickness of the erodible layer was $h_{stop}$. 425 $\pm$ 0.9 g of 2 phi particles were then released on this surface from a 10.5-cm diameter cylindrical container. The propagation of the sand was measured by videotaping while a horizontal grid was projected onto the plane to measure flow thickness. Because of geometrical distortions and the difficulty in ascertaining the edge of the flow during periods when the material was thinly spread, the error in the measurements of the flows positions is estimated to range from 1 -- 2.5 cm. The sand was allowed to flow off the end of the inclined plane into a bin. We separated the two fractions by sieving. Using this technique, we were almost always able to ensure that $< 1%$ of the blue, 2 phi particles were lost due to experimental error. It was possible to determine an erosion gradient for the experiments by starting separarate runs at each angle at different positions on the inclined plane. In this way, particles were collected after flowing different distances down the plane. Using the erosion gradient and flow speed data, we estimated erosion rate. It was possible to quantitatively match data on amount of material eroded and flow speed with the numerical model. Because it was not possible to distinguish bed particles in the numerical model, the results were summarized by comparing the mass of particles transported off the end of the slope. The results show that as slope angle increased and distance of the starting mass from the end of the slope decreased, the total mass of particles transported off the slope increased in both data and model. The results suggest that the numerical model is able to reproduce erosion conditions reasonably well over a range of slope angle and slope length conditions. The numerical model could be used to understand the process and effects of erosion in extreme geophysical granular flows.
DE: 1815 Erosion and sedimentation
DE: 3210 Modeling
DE: 3220 Nonlinear dynamics
DE: 8499 General or miscellaneous
SC: Hydrology [H]
MN: 2003 Fall Meeting