Loading…

APPLICATION OF A SIMPLIFIED BOUSSINESQ DEPTH AVERAGED MODEL TO NATURAL COMPLEX RIVERS

Natural complex river reaches often represent some of the most difficult topographic and hydrodynamic phenomena to evaluate and simulate in detail. A clear understanding and an accurate representation of the hydraulic characteristics of such rivers are necessary. Numerical models offer the potential...

Full description

Saved in:
Bibliographic Details
Published in:Journal of engineering and applied science (Online) 2004-06, Vol.51 (3), p.523-542
Main Author: Ghamry, H K
Format: Article
Language:English
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Natural complex river reaches often represent some of the most difficult topographic and hydrodynamic phenomena to evaluate and simulate in detail. A clear understanding and an accurate representation of the hydraulic characteristics of such rivers are necessary. Numerical models offer the potential for analyzing the quality and quantity of such challenging circumstances. A two-dimensional simplified Boussinesq depth averaged model was proposed in this study. The model was derived having integrated the three-dimensional Reynolds-averaged Navier-Stokes equations. The proposed model was applied to natural river flow cases to check its validity to simulate such real problems with complex configuration. These problems included flow in rivers with bifurcation and islands and with different degrees of meandering. Typical study sites on the Assiniboine and Saskatchewan Rivers, Canada were investigated. The implicit Dissipative Galerkin finite element scheme was used in these simulations. Computed values for longitudinal water surface profiles and water surface across the river widths were compared to detailed representative field measurements. A good agreement was attained. Moreover, this study showed that the proposed model is robust and accurate for simulating such complex natural rivers. This should be true for small-scale models where the generated numerical meshes are very fine.
ISSN:1110-1903