Date of Award


Document Type

Campus Access Dissertation


Civil and Environmental Engineering

First Advisor

Jasim Imran


The work presented in this dissertation consists of experimental and numerical studies on dam-break and levee-breach flows. Experiments were conducted to gather comprehensive data on a steady levee-breach flow in an urban area and on unsteady 2-D and 3-D dam-break flows. Numerical studies of dam-break flows were carried out using Reynolds-averaged Navier-Stokes (RANS) approach and large eddy simulation (LES) modeling for turbulence. The volume of fluid (VOF) was used for tracking the free surface.

To study levee-breach flooding in an urban area, laboratory experiments were conducted on a 1:50 scale model of the 2005 17th Street Canal levee-breach in New Orleans, Louisiana. Steady state flow depths and flow velocities were measured by a point gauge and an array of ultrasonic velocity profilers (UVP), respectively. Surface velocities were recorded in the entire model using a digital particle tracking velocimetry (DPTV) technique. The results of this investigation showed that closely spaced buildings acted like a single obstacle with a common wake zone, with very little flow between these buildings. The results also showed that the strongest flow was in the pre-breach direction and the local topography played an important role in distributing the flow in the flooded area. The complete data set may be utilized to validate shallow water models and 3-D computational fluid dynamics models of free-surfaceflow.

A two-dimensional idealized dam-break experiment was conducted in a straight flume. Dam failure was modeled by suddenly lifting a gate inside a flume for different upstream reservoir levels. Ultrasonic Doppler Velocity profilers (UVP) recorded transient velocity profiles at eight different locations upstream and downstream of the removed gate. This experiment provided data on the spatio-temporal evolution of the flow field in an unsteady flow of relatively short duration. The 2-D experiment was simulated using a computational fluid dynamics (CFD) solver, FLUENT. The UVP probes captured the transient velocity profiles. Based on the results, the following observations were made; (i) turbulence modeling does not affect the velocity profile in the upstream reservoir, but has significant influence on the downstream velocity; (ii) the velocity magnitude at a specific location changes with time, but the shape of the velocity profiles remain similar; (iii) an analytical solution for frictionless dam-break flow on a sloping bed and numerical simulation using LES modeling show satisfactory agreement with measured water surface profile; (iv) non-dimensionalization of reservoir-side velocity profiles resulting from different reservoir heads and at different locations from a specific head show a collapse and reveal that the shear layer thickness of these profiles is approximately 5% of the initial reservoir head; (v) measurement and simulation with the large eddy simulation (LES) model for turbulence show satisfactory agreement, suggesting that the LES modeling is a viable approach for an accurate prediction of dam-break flows.

Three-dimensional numerical and experimental studies were conducted to analyze partial-breach dam-break flows. Simulations were carried out using ANSYS Fluent. Results were compared with published data on dam-break experiments and simulations carried out by others using shallow water equations (SWE) modeling. The results showed that both the LES and the k -epsilon; modeling satisfactorily reproduced the temporal variation of the measured bottom pressure, however, the LES model captured the free surface and velocity variation with time more accurately.

To gather additional data using recently developed measurement techniques, a partial breach dam-break experiment was conducted in a large setup in the Hydraulics Laboratory. Measurements of particle tracking velocimetry (PTV), UVPs and pressure sensors, 3-D surface velocity, velocity profiles in the reservoir, and static and dynamic pressures were obtained from this experiment. The dam-break experiment was repeatable and symmetric. Based on the results, the following observations are made; (i) the near-field region of the dam-break flow is characterized by a 3-D flow having the same order of magnitude for the vertical and the streamwise velocity components (ii) Comparison of 3-D surface velocity data between measurements and simulations showed satisfactory agreement; (iii) Hydrodynamic pressure increased with the distance away from the dam; (iv) A complete velocity profile measured in the upstream reservoir near the breach showed that velocity decreased towards the water surface; (v) There was excellent agreement between measured and simulated flow fields including pressures, water surfaces, surface velocities, and velocity profiles.