Graded Sediments

Sediments can be divided into many size ranges so that the movement of graded sediments can be adequately represented. A representative size is input for each sediment fraction. The sediment transport rate for each fraction is computed by multiplying the predicted rate (from the sediment transport equation) by the proportion of the bed material which consists of that fraction. The bed material composition is either assumed to be constant and is set at input or it is predicted from the rates of deposition or erosion at each section. The former approach is termed the 'COMPOSITE' algorithm, as a single 'composite' grading represents the channel bed material. The latter approach is termed the 'SORTED' algorithm because sediment sorting effects are modelled.

Under the SORTED algorithm, the composition of three layers is recorded: an active layer at the surface of the channel bed, a deposited layer below it and the parent bed material below the deposited layer (see Figure 1). The quantity of sediment in each of these layers is recorded at each channel section for each sediment fraction. If there is net erosion at a section, then the deposited layer is not present. As deposition or erosion causes an exchange between the material in the active layer and the material in transport, the composition of the active layer is updated. For example, in the case of erosion from a widely graded channel bed by clear water, the finer sediments will be brought into transport more rapidly than the coarser sediments. So the proportion of the active layer consisting of the coarsest fraction will increase and the proportion of the finer fractions will decrease. If some fractions are not eroded at all, then the active layer will eventually stabilise to an armoured state for which only non-mobile sediments are present in the active layer. The active layer thickness is set at input and does not change so that for erosion the material eroded from the active layer is replaced by an equal quantity of material taken from the layer below (the deposited layer if present or the parent material otherwise). In the case of deposition, the material being added into the active layer is matched by an equal volume of material passed from that layer into the deposited layer.

If the sediments are fairly uniform, say D/D< 2.0, then a single sediment size can be used and no sorting effects need be modelled. In this case the COMPOSITE algorithm should be used. In other cases the SORTED algorithm is recommended (Figure 1 ). Significant errors can occur if the COMPOSITE algorithm is used in situations in which the bed composition would vary through time.

An important feature of sediment transport when the bed material is widely graded is the marked difference in sediment composition between the bed material and the material in transport. The material in transport will be considerably finer; in some cases the median size can be finer by several orders of magnitude. 1D Sediment Transport Solver has input facilities that allow a difference in grading between the sediment inflow and the bed material at the inflow point. Advice on using these facilities is given in the final section of this manual.

The bed material input to the model can be made to vary from section to section when the COMPOSITE algorithm is used. If the SORTED algorithm is used these facilities enable the parent bed material to be varied from section to section (the parent bed material is also used to set the initial composition of the active layer). See section Datafile Format.