Doctor of Philosophy (PhD)


Civil and Environmental Engineering

Document Type



The Lower Mississippi River Physical Model (LMRPM) is a distorted-scale, 1:6000 horizontal and 1:400 vertical, movable-bed model that is being used to complement ongoing numerical and field studies directed at studying how non-cohesive sediment transport will be impacted by future changes in relative sea level rise and river discharges, as well various management strategies, such as river sediment diversions, in the lower ~190 miles of the Mississippi River The LMRPM was designed using Froude number (Fr) similarity between the prototype and the model, while the flow Reynolds number is relaxed, but determined to be high enough to ensure rough turbulent conditions. There are, however, questions about whether the Reynolds number approach, used in the design, ensures fully turbulent flow.

It is well known that scale effects due to model distortion arise due to differences in force ratios between the model and its real-world prototype. Model distortion cause differences in the centrifugal forces, skewed helical flow patterns through bends, and steeper channel bank walls.

To investigate the impact of model distortion on the two-dimensional flow fields, two numerical models were developed using MIKE 21C software: one at the prototype scale and one at the LMRPM scale. The river curvature was calculated for all cross-sections in the model domain and used to classify each cross-section as straight, moderate or sharp. Both models were calibrated and validated, and comparisons, over a range of river discharges and reach types, were made between several two-dimensional velocity fields properties. In addition, the Dean Number, a dimensionless parameter that includes the river curvature, was calculated, and used to look at the levels of turbulence.

The two-dimensional flows in the straight and moderate bend reaches were mostly the same over the range of river discharges. However, the flow fields in the LMRPM sharp bends showed no or smaller separation zones, when compared to the prototype flow fields, due to the lower relative centrifugal forces. The walls steepness has a significant impact on the flow fields by pushing the water to the centerline of the channel and preventing deviation of the velocity vectors. The deviations from prototype flow fields also created different patterns in the eddy viscosity and bed shear stress values. While both the vorticity and Dean Number values at each cross section are indirect measures of some processes that can promote turbulence, both indicate that there are hydrodynamic processes, even in areas where Reynolds numbers might only indicate transitional flow, that should contribute to the generation of more turbulence. These results indicate that the LMRPM design does generate sufficient turbulence at the medium and high river discharges when the non-cohesive sediment transport is occurring.



Committee Chair

Willson, Clinton