Search

A model is proposed to determine the sediment transport rate generated by the combined action of sea waves and tidal currents, the latter being modelled as a sequence of steady currents because of the large values of the Keulegan-Carpenter number of the tidal flows, which allow to neglect inertial effects in their modelling. The hydrodynamic module solves Reynolds-averaged momentum equations by introducing a two-equation turbulence model, which considers also the flow in the region closest to the sea bottom and enforces the no-slip condition and the vanishing of the turbulent kinetic energy at the bottom without the need to assume the existence of a logarithmic velocity profile and to fix the normal derivative of the turbulent kinetic energy close to the sea bed. Hence, the evaluation of the velocity profile in the near-bed region turns out to be accurate. The sediment transport module determines the sediment concentration and sediment transport for a mixture of non-cohesive and cohesive sediments. To validate the model, its predictions are compared with laboratory measurements and field data collected over rippled beds at two different tidal estuaries. Then, results are obtained for values of the parameters chosen to mimic a site close to Punta Umbría (Huelva, Spain) and considering different hydrodynamic conditions and different sediment mixtures. The influence of the waves on the suspended sediment concentration is lower for the cohesive fraction than for the non-cohesive fraction. For all the sediment mixtures, it turns out that the cohesive fraction is well mixed over the whole water depth, meanwhile the concentration of the non-cohesive fraction has significant values only near the bottom, thus showing that an accurate description of turbulence dynamics in the buffer layer and viscous sublayer (if it exists) is important to quantify the sediment transport rate. • Sediment transport module that reproduces the concentration of cohesive fraction is presented. • Tidal current is modelled as a sequence of steady-currents due to the large values of the Keulegan-Carpenter number. • Menter's turbulence model to predict the turbulence structure both in the wave bottom boundary layer and in the core region. • Turbulence dynamics description in the buffer layer and viscous sublayer is important to quantify the sediment transport. • Applicable to any type of sediment mixture. [ABSTRACT FROM AUTHOR]

In the present article two flocculation models, given in [5] (LG model) and [36] (S model) are compared. Both models give the time evolution d L / d t where L is the size of a particle undergoing flocculation, and t is the time. The LG model is based on logistic growth theory, whereas the S model is based on the theory originally derived by Smoluchowski. Both models have the advantage of easy implementation in, for instance, large-scale sediment transport numerical models. However, it is found that they do not obey the same kinetics. A series of laboratory experiments is presented where the flocculation of a mineral clay by polyelectrolyte is studied as a function of clay concentration and shear rate. From modelling these experiments, it is found that the LG model reproduces the time dependence of the floc sizes found experimentally, whereas the S model does not. It is shown that the LG model can be used to model the data over the whole range of clay concentration and shear investigated. Based on the study presented in this article, it was found that the average floc growth in time for the clay type and conditions applied in the experiments could be modelled by: d L / d t = 40 × 1 0 − 4 G 0. 75 × [ e x p (− 2 × 1 0 − 4 G 0. 75 t) ] / [ 1 + 20 e x p (− 2 × 1 0 − 4 G 0. 75 t) ] L. • Logistic growth theory is compared with Smoluchowski-based theory. • The models do not have the same kinetics. • Only logistic growth theory agrees with experiments. • Logistic growth theory models data for all clay concentrations. • Logistic growth theory models data for all shears investigated. [ABSTRACT FROM AUTHOR]

Abstract: Sediment cores from the western Gulf of Lions France were subject to known bottom shear stresses with the goal of understanding size-specific sediment erodibility. On cruises in October 2004, February and April 2005, cores with an undisturbed sediment–water interface were collected along a transect extending seaward from the Tet river mouth. The cores were exposed to increasing shear stresses (0.01–0.4Pa) onboard the vessel shortly after collection by using a Gust erosion chamber. Samples of the suspensate were collected during the erosion experiments and analyzed for disaggregated inorganic grain size (DIGS) using a Coulter Multisizer IIe. Size-specific mobility plots were generated by dividing the proportion of each grain size in suspension at each shear stress by its proportion in the sediment before erosion. If all grain sizes that make up the bottom sediment are eroded equally from the bed, then mobility equals one for all grain sizes. Values >1 indicate that the suspended sediment is enriched in the size class and values <1 indicate that the size class is enriched in the bed. Results show that in non-cohesive, sandy silts, fine grains (clays and fine silts) are eroded preferentially from the bed at low shear stresses. With increasing bottom stress progressively larger grains are eroded from the bed. In cohesive silts, preferential erosion of the finer sizes no longer occurs, with all sizes up to medium silts eroding at approximately the same rate. Effectively, a sandy silt can be winnowed of its fine grain fraction during erosion while cohesive silts cannot. This difference in the sortability of cohesive and non-cohesive sediment during erosion may control the position and maintenance of the sand–mud transition and the sequestration of surface-adsorbed contaminants. [Copyright &y& Elsevier]

Abstract: An important element within the Defra funded Estuary Process Research project “EstProc” was the implementation of the new or refined algorithms, produced under EstProc, into cohesive sediment numerical models. The implementation stage was important as any extension in the understanding of estuarine processes from EstProc was required to be suitable for dissemination into the wider research community with a level of robustness for general applications demonstrated. This report describes work undertaken to implement the new Manning Floc Settling Velocity Model, developed during EstProc. All Manning component algorithms could be combined to provide estimates of mass settling flux. The algorithms are initially assessed in a number of 1-D scenarios, where the Manning model output is compared against both real observations and the output from alternative settling parameterisations. The Manning model is then implemented into a fully 3-D computational model (TELEMAC3D) of estuarine hydraulics and sediment transport of the Lower Thames estuary. The 3-D model results with the Manning algorithm included were compared to runs with a constant settling velocity of 0.5mms−1 and settling velocity based on a simple linear multiplier of concentration and with the above mentioned observations of suspended concentration. The findings of the 1-D case studies found the Manning empirical settling model could reproduce 93% of the total mass settling flux observed over a spring tidal cycle. The floc model fit was even better within the turbidity maximum (TM) zone. A constant 0.5mms−1 only estimated 15% of the TM mass flux, whereas the fixed 5mms−1 settling rate over-predicted the TM mass flux by 47%. Both settling velocity as a simple linear function of concentration, and van Leussen''s method, did not fare much better estimating less than half the observed flux during the various tidal and sub-tidal cycle periods. When the Manning-settling model was applied to a layer with suspended concentrations approaching 6gl−1, it calculated 96% of the observed mass flux. The main conclusions of the implementation exercise were that it was feasible to implement a complex relationship between settling velocity and concentration in a 3-D computational model of estuarine hydraulics, without producing any significant increase in model run times or reducing model stability. The use of the Manning algorithm greatly improved the reproduction of the observed distribution of suspended concentration both in the vertical and horizontal directions compared to the other simulations. During the 1-D assessments, the Manning-settling model demonstrated flexibility in adapting to a wide range of estuarine environmental conditions (i.e. shear stress and concentration), specifically for applied modelling purposes. [Copyright &y& Elsevier]