Three-dimensional modelling of secondary flow in river bends

Three-dimensional modelling of secondary flow in river bends

Jongbloed, J.W.

Stelling, G.S. (mentor)
Battjes, J.A. (mentor)
Di Silvio, G. (mentor)
Wang, Z.B. (mentor)
Booij, R. (mentor)

Civil Engineering and Geosciences

Hydraulic Engineering

1996-10-01

The flow in a river bend has a large influence on the cross-sectional profile of the bend. Due to the curvature of the bend a secondary flow, which is perpendicular to the main flow, occurs. The secondary flow is directed outwards in the upper part of the cross-section and inwards in the lower part of the cross-section. It causes by means of the transverse transport of main flow momentum a redistribution of the main flow. This redistribution of the main flow and the sediment transport by the secondary flow cause a typical river bend profile, a steep sloping bank in correspondence with a large depth near the outer wall and a smoothly sloping bank in correspondence with a small depth near the inner wall. To be able to predict (the changes of) the profile and the position of a river bend one must understand the flow pattern and the related sediment transport. To account for the flow pattern several numerical simulation programs have been developed. Since, nowadays, it is impossible to calculate the flow exactly, due to the turbulence, it has to be modelled and some assumptions and approximations have to be made. At Delft Hydraulics a program package (Delft3D) has been developed to simulate these flow cases and their sediment transports. In this thesis research has been done to what extent Trisula, the part that accounts for the prediction of the fluid movement, is capable of predicting the flow in river bends. At first a description of the flow pattern in river bends and a summary of the possibilities of the turbulence modelling are given as well as an abstract of the previous research, both numerically and experimentally, to gain insight in the phenomenon. To verify Trisula an experiment in an 'infinite river bend' has been done at the University of Padua at the Institute of Hydraulics 'G Poleni'. Due to the sensitivity of the boundary conditions this experiment gave not satisfying results so the program had to be checked with other measurement. The measurements of De Vriend were used to verify the computational results of a strongly curved bend and the measurements of Booij were used to verify the computational results of a smoothly curved bend. From the research to the simulation of the flow in river bends, it appeared that Trisula is able to predict the main features that occur in a river bend although the magnitude, especially of the radial velocities, is sometimes too small. The velocity distributions over the vertical and the development and decay of the secondary flow throughout a river bend are predicted rather well. One of the most striking features is the impossibility of trisula to predict the counterrotating secondary flow near the outer wall at sufficiently large Dean numbers due to the incorrect modelling of the turbulence

river morphology
river bend
secondary flow

http://resolver.tudelft.nl/uuid:d67d1c10-cde9-498e-8e0a-e38d3634d787

Student theses

master thesis

(c) 1996 Jongbloed, J.W.