Siltation of harbours and coastal zones is often undesirable as it diminishes navigability. Besides, deposited mud is often polluted, because it specifically adsorbs contaminants from the water. For managers of harbours and coastal zones it is therefore important to know prior to execution whether planned projects will diminish or increase siltation. The performance of the current cohesive sediment transport models, which are being used as a tool for making predictions, is sometimes disappointing and therefore needs further improvement. The shortcomings are caused to a great extent by an insufficient knowledge of mechanisms underlying cohesive sediment transport.
Of these mechanisms some are studied in this thesis, particularly the generation and transport of concentrated fluid mud layers near the sea bed. For this purpose, first the material properties such as strength and viscosity of mud both in fluid form and solid form are determined, using rheological instruments and a miniature sounding probe. Subsequently, the liquefaction of deposited mud by waves is studied with laboratory experiments, as well as the resulting flow on a sloping bed caused by gravity. Mathematical models aiming at describing these phenomena are compared with the experiments. Finally, these models are used to evaluate the behaviour of mud under natural conditions. In support of this, also some field observations reported in the literature are analyzed.
The conclusion is that mud layers that have been deposited under quiet conditions may suddenly become liquid under storm conditions and can subsequently be transported. The likelihood of this type of behaviour depends on the strength of the deposits, which is related to the degree of consolidation after deposition. Transport of fluid mud layers may result in sudden changes in bathymetry. Especially in navigation channels this phenomenon may occur, as flow of fluid mud on adjacent slopes is enhanced by gravity. This should be well taken into account when modelling. Rheological properties such as strength and viscosity should not be assumed to be constant, but should be set to be dependent on the actual stress condition and the stress history.