Three-dimensional modelling of a very large river: the Rio Paraná

S.D. Sandbach, R.J. Hardy, S.N. Lane, D.R. Parsons, J.L. Best, Philip Ashworth, A.J.H. Reesink, M.L. Amsler, R.N. Szupiany, A.P. Nicholas, O. Orfeo, G.H. Sambrook-Smith

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNChapterpeer-review

Abstract

Our current understanding of the world’s largest rivers is limited due to the difficulties in obtaining detailed, spatially-distributed, flow and morphological data. Recent progress has been made in obtaining data-rich results, using Computational Fluid Dynamic (CFD) models within smaller river reaches, but the application of CFD to large rivers has not been fully developed due to our inability to: i) design numerically-stable meshes for complex topographies, and ii) collect high resolution data appropriate for model boundary conditions. Here, we test a new five-term mass-flux scaling algorithm to include bed topography over large spatial scales, where the discretised form of the mass and momentum equations are modified using a mass-flux scalling approach. Converged solutions were obtained using Reynolds-Averaged Navier Stokes (RANS) equations with a κ–ε RNG turbulence closure. Topographic and velocity data were collected for a 38 km reach in the Río Paraná, Argentina, to provide boundary conditions and validation data. The model results demonstrate the importance of topographic steering on determining flow structures in large rivers and the difficulties in objectively identifying coherent secondary flows.
Original languageEnglish
Title of host publicationProceedings of the IAHR River Flow Conference
EditorsA. Dittrich, Ka Koll, J. Aberle, P. Geisenhainer
Place of PublicationGermany
PublisherBundesanstalt fuer Wasserbau
Pages409-417
Number of pages9
Volume1
ISBN (Print)9783939230007
Publication statusPublished - 1 Sept 2010

Fingerprint

Dive into the research topics of 'Three-dimensional modelling of a very large river: the Rio Paraná'. Together they form a unique fingerprint.

Cite this