Testing a Multiple Grain-Class Mud Deposition Algorithm

A numerical algorithm intended for the depositional phase of a dynamic multiple-grain-class mud-sediment transport model is presented. The algorithm is based on the paradigm of mutually exclusive erosion and deposition, stress-dependent fractional deposition with a lower stress limit leading to complete deposition, and cohesion-limited sorting/segregation of grain classes. As the algorithm's basis, the analytic grain size model of Kranck and Milligan was modified to eliminate concentration- in favor of stress-dependence and to cast it in terms of sediment fraction instead of absolute concentration. The algorithm establishes a dynamic threshold for deposition based on the 95 percentile finer-than size class of the suspension and deposits material (or not) depending on the bed shear stress. At some lower stress threshold the algorithm switches and allows complete deposition at a rate dependent on a Krone's depositional probability. During deposition, grain classes are coupled (as by cohesion) so that the size distribution decreases most rapidly at its upper limit and by a certain proportion over the remaining classes to maintain an objective self-similar size distribution. Example calculations are presented based on particle size and concentration data from previous annular-flume deposition tests.