Transverse slope effects on widely graded sediment: Experimental and computational research

Transverse slope effects on widely graded sediment: Experimental and computational research

Schoonen, P.E.M.

de Vriend, H.J. (mentor)
Fontijn, H.L. (mentor)
Mosselman, E. (mentor)
Sloff, C.J. (mentor)

Civil Engineering and Geosciences

2006-02-01

The floods of the River Rhine and the River Meuse in 1993 and 1995 pointed out the importance of river intervention works to maintain safety for surrounding inhabitants. To predict the consequences of such engineering measures on river morphology, numerical computer models are used. Due to the high socio-economic value of the Dutch river system and the importance of continuous navigability, possible disturbances in this system have to be carefully analysed. To gain a higher accuracy, the gradation of sediment has recently been implemented in these models. An important factor in river bed morphology is the gravitational influence on the direction of sediment transport e.g. in river bends and on side banks. This influence is modelled using deterministic formulae for uniform sediment. The main objective of this study is to review the applicability to widely graded sediment of the present formulae for transverse slope effects. This is done by comparing the measurements of a laboratory experiment to an analytical model and to the output of numerical simulations. The laboratory research that was conducted at the Fluid Mechanics Laboratory of Delft University of Technology, consisted of four levelling experiments at different flow velocities and with different sediment compositions. Three experiments were done using a tri-modal sediment mixture and one experiment with a uniform mixture. A transversely sloping profile that is derived from the analytical model, was constructed from the sand mixture in a laboratory flume. A uniform flow was then created and the movement of the bed was measured as it slowly levelled towards a cross-sectional horizontal equilibrium. The analytical model is a first order approximation of a small transverse distortion, derived from the continuity and transport equations using the deterministic formula of Talmon for the transverse slope effect. As this model is representing only uniform sediment, a numerical simulation in Delft3D, using graded sediment calculations, is added using the formulae of Koch and Flokstra and of Parker and Andrews for transverse slope effects. When comparing the development of the measured transverse profiles to the models, a good agreement is found for the experiment with uniform sediment, but a constant underestimation of the involved timescales is found for the tri-modal mixture. In order to find a better agreement of the numerical model, the parameters of the Koch and Flokstra formula for transverse slope effect were varied to fit the timescales of the simulations for the tri-modal mixture to that of the experiments. Considering the design and the execution of the levelling experiments it can be concluded from this study that the method used is adequate. Considering the morphological modelling of transverse slope effects it can be concluded that none of the tested formulae for transverse slope effect can accurately predict the transverse deviation of sediment transport within their default parameter range.

graded sediment
river
Rhine
Meuse
transverse profiles

http://resolver.tudelft.nl/uuid:b3afd6eb-6db4-4506-aebc-42447aaef756

TU Delft, Civil Engineering and Geosciences, Hydraulic Engineering

Student theses

master thesis

(c) 2006 P.E.M. Schoonen