Net sediment transport under sheet flow conditions

Abstract

This study focuses on time-averaged sediment transport in oscillatory flow in the sheet flow regime. The tests, used in this study, are part ofthe experiments, carried out under the name SISTEX99 (Small-scale International Sediment Transport Experiments 1999). SISTEX99 focuses on near-bed sediment transport processes under influence of regular and irregular non-breaking waves in different wave conditions. The main goal is to increase insight in these processes by obtaining quantitative data of bedform characteristics, near-bed flow velocities, near-bed sediment concentrations and net transport rates. This information can be used for the verification, improvement and development of mathematical models. The main objectives of this study were: 1) To obtain a detailed data set on time-averaged sediment transport rates, measured under progressive waves in sheet flow conditions. 2) To compare the measured time-averaged sediment transport rate with different sediment transport models and other experimental data sets in order to verify if there are differences between net transport rates, measured in a purely horizontal oscillatory water motion (water tunnel) or measured under progressive waves. Experimental research The experiments for this study were carried out in the Large Wave Flume (Grosser WellenKanal, GWK) of the ForschungsZentrum Kiiste (FZK) in Hannover, Germany. Four different test conditions in the sheet flow regime with monochromatic asymmetrical waves with varying wave heights and periods, above a horizontal sand bed, were carried out. The sand bed consisted of uniform sand with a median grain diameter of 0.24 mm. The net sediment transport rates were calculated from the measured bed profiles, using a mass conservation technique. Near-bed oscillatory flow velocities were measured, just outside the wave boundary layer. Also spatial measurements of the wave heights and periods were carried out during the tests. Sediment transport models The net transport rate data are used for the verification of four sediment transport models: three quasi-steady models (Bailard, 1981; Ribberink & AI-Salem, 1993; Ribberink, 1998) and one semi-unsteady model (Dibajnia & Watanabe, 1992). All transport models underpredict the measured net transport rates. The quasi-steady model of Bailard (1981) underpredicts the measured net transport rates with about a factor 2. The quasi-steady model of Al-Salem (1993) underpredicts the measured transport rates. When coefficient A=5 (1993) is applied, the measured transport rate is underpredicted with about a factor 3 and applying coefficient A=4 (1994) leads to an underprediction with about a factor 3-4. The quasi-steady model of Ribberink (1998) underpredicts the measured net transport rate with about a factor 4. The semi-unsteady model of Dibajnia & Watanabe performs best. The transport rates are underpredicted with about factor 1-2. Other experimental data-sets In addition, a comparison is made with three other experimental data sets, obtained from the Large Oscillating Water Tunnel (LOWT) of Delft Hydraulics: series B, conducted by AISalem (1993), series E, conducted by Katopodi et at. (1994) and series J, conducted by Janssen & V.d. Hout (1997). The experimental conditions of Al-Salern consisted of regular and irregular asymmetric oscillatory water motions. The conditions of Katopodi et at. (1994) and Janssen & V.d. Hout (1997) consisted of a symmetrical oscillatory water motion, superimposed on a net current. In series B, E and J sand with a median grain diameter of 0.21 mm was applied, opposed to 0.24 mm of the present experimental sand. A comparison with the experimental data-set of Al-Salem (1993) shows that the present results are about a factor 2-3 larger than results under comparable conditions. The present results are larger than series E, conducted by Katopodi (1994) and series J, conducted by Janssen & V.d. Hout (1997), larger than about a factor 7-8.