Dynamics of Unconfined Turbidity Currents and Their Interaction with Obstacles and Substrates

Abstract

Turbidity currents are a sediment transport process of importance to water resources engineers. Typically initiated by rivers in flood or failure of sedimentary slopes, they threaten sub-aqueous infrastructure – including the ability to rupture submarine cables. They are also a significant driver behind reservoir sedimentation, where recent laboratory-based research investigated the use of obstacles to reduce current propagation. These studies have focused on confined flume environments. This thesis aims to deepen our understanding behind obstructed turbidity currents propagating over various substrates in unconfined environments – focusing on ambient fluid entrainment mechanisms and the influence of lateral unconfinement. To achieve this, a range of scaled laboratory experiments were undertaken on a turbidity current interacting with a rectangular obstacle, in both confined and unconfined environments. A secondary aim is to introduce a novel photometric technique that can be synchronised with conventional flow measurement instrumentation, providing qualitative and quantitative insight into current development. The developed technique comprises a spatial calibration and light intensity-thresholding process, with additional noise-reducing algorithms. Successful integration of the measuring techniques allows high spatio-temporal resolution images of current development to be captured in addition to internal density and flow structure. This provides visual insight into the mixing instabilities, along with quantitative insight into spatial development and entrainment parameterisation. Four stages of ambient fluid entrainment are observed as the current traversed the obstacle: (i) lateral entrainment stage; (i) jet stage; (iii) collapsing stage; and (iv) reestablishment stage. The Morton-Taylor-Turner hypothesis was confirmed for both confinement cases, however only unconfined tests show a relationship with Reynolds, Froude and Richardson numbers. The role of the substrate is found to be insignificant on ambient fluid entrainment rate, contrary to previous studies. Although confined and unconfined current entrainment and internal flow structure are found to be similar in the stream-wise direction, the unconfined currents decrease in density more rapidly and have a lower boundary height. Therefore, confined laboratory models may over-estimate the obstacle height needed to block current propagation.