240 likes | 264 Views
Computing Flow and Sediment Transport over Bedforms. Jonathan Nelson Aaron Burman Yasuyuki Shimizu Steve McLean Ron Shreve Mark Schmeeckle. Department of the Interior U.S. Geological Survey. Objective.
E N D
Computing Flow and Sediment Transport over Bedforms Jonathan Nelson Aaron Burman Yasuyuki Shimizu Steve McLean Ron Shreve Mark Schmeeckle Department of the Interior U.S. Geological Survey
Objective • A practical computational morphodynamics model for predicting the behavior of bedforms • Predict geometry of bedforms accurately • Predict pressure with sufficient accuracy to determine form drag • Suitable for initiation, development to equilibrium states in steady flows • Suitable for predicting adjustment in time-varying flows
Morphodynamic Modeling • Flow Model • Sediment-transport Model • Exner equation for morphologic change
Flow Model • Large-eddy simulation, solves the Navier Stokes equations with subgrid-scale closure,CIP technique • Unsteady two-dimensional model with nonlinear k-epsilon closure
Instantaneous streamwise velocity at the bed Both graphs are vertically exaggerated (real scale: dunes are 4cm tall, 90 cm wide and 80cm long) Time-averaged and vertically-averaged streamwise velocity (50 sec)
Large-eddy Model • Works well for two-d dunes for both velocity and pressure • Works almost as well more complex three-d dunes • Takes too long for morphodynamics modeling of realistic geometry Need to use a closure model
Sediment-transport Model • Direct simulation of particle motion using model developed by Schmeeckle • Pickup/deposit functions from Nakagawa and Tsujimoto
Development of bedforms from an initially perturbed bed- Sinusoidal perturbation,10-cm wavelength, 1-cm height
Conclusions • Large-eddy simulation models are the future of computational morphodynamics, but we’re not quite there in terms of computational speed. • Simpler closure type models are not as accurate, but appear to be accurate enough for bedform morphodynamics. • Experiments and numerical results strongly suggest that local flow variability is critically important for predicting transport in complex flows. • Next step is temporal response to varying flows, bar/bedform interactions, intermediate features.