Use this URL to cite or link to this record in EThOS:
Title: Numerical modelling of hydraulic free surface flows and scale effects associated with physical modelling
Author: Torres Mansilla, Caterina
ISNI:       0000 0004 7960 2919
Awarding Body: University of Leeds
Current Institution: University of Leeds
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
The recent increase in frequency and severity of extreme weather events has resulted in the crucial need for design and rehabilitation of hydraulic infrastructure like weirs and spillways. These structures play an essential role by ensuring flood protection and security of water resources. The current scenario has triggered an increased implementation of labyrinth weirs, which enable greater efficiency where larger discharges are expected. The standard means of hydraulic modelling for the design of this type of hydraulic structures consist in scaled physical hydraulic models. The principal limitation of these experimental techniques is their associated scale effects which are induced by the impossibility to equate all force ratios in the prototype and model. Renewed research is needed in order to determine whether such distortions are present in physical models of labyrinth weirs and provide refined limits to minimise them. Moreover, in the recent years, interest in numerical modelling has grown amongst the hydraulic structures community. Several Computational Fluid Dynamics (CFD) techniques have been proposed for hydraulic modelling, enhanced by dramatic improvements in computer processing power. These approaches require further validation evidence for a wider range of structures and flow conditions to demonstrate their reliability and to inform best practice on their implementation. Determination of the extent to which the leading numerical approaches are capable of reproducing an experimental flow of interest is therefore of significant importance. The present work includes the initial evaluation of the capability of two leading numerical techniques to reproduce an experimental free surface flow and focuses on the assessment of the Volume of Fluid (VOF) method to simulate the flow over a labyrinth weir and investigate scale effects of a physical model. The Smoothed Particle Hydrodynamics (SPH) technique and the VOF method are first tested for a dam break case over an obstacle. Subsequently, the VOF is employed in two solvers (ANSYS Fluent and OpenFOAM) to simulate the physical model of a labyrinth weir and spillway. After validation in respect of various flow aspects is undertaken, the prototype scale is simulated, and scale effects are examined. Finally, limits to minimise scale effects observed in the different flow aspects (depths and velocities in the spillway channel as well as in the labyrinth weir rating curve) are estimated based on the numerical predictions. The VOF modelling of various flow aspects in the physical hydraulic model demonstrated the RANS k - ε family models implemented with the PLIC interface capturing scheme were appropriate to characterise the flows encountered over the labyrinth weir and in the spillway channel. In order to achieve mesh independence, the VOF applied in Fluent required a minimum cell size of 8x10-3 m and in OpenFOAM required 4x10- 3 m. For the lowest flow rates, the minimal discrepancies observed in the predictions from the two solvers were found to be due to the interface capturing scheme. For the largest flows, more significant differences were found between the two solvers which were due to cell size sensitivity. This study demonstrated that the 3D CFD VOF with the appropriately chosen numerical implementations is capable of reproducing the complex free surface flows over and downstream a labyrinth weir for a range of flow conditions. The comparison of the prototype and physical model scale VOF predictions revealed the occurrence of larger velocities and lower depths at prototype scale. The differences at the two scales were manifested in the spillway channel flows as well as in the weir rating curve and decreased for increasing flow rate. Prototype scale simulations also showed increases in the weir nappe, causing elongation of the cross-wave configuration generated by the labyrinth weir. These were found likely to be caused by differences in pressure distribution at the weir crest and were reduced for increasing flow rate. The above findings were very well correlated with existing experimental studies from the literature. In addition, the prototype scale simulations presented changes in the waves' positions, occurring even for the largest flow rates where the scale effects on depth and velocity were minimal. Simulation results of additional scales 1:50 and 1:10 indicated that the waves' displacements are reduced for decreasing scale factor of the simulation. Limits to minimise scale effects observed in the labyrinth weir rating curve as well as in the depths and velocities in the spillway channel were estimated using the Fluent numerical predictions. The derived limits were in close agreement with existing limiting criteria found in the literature. The present work substantiates the capability of CFD as a technique to quantify scale effects induced by physical models and determine limits to minimise them.
Supervisor: Borman, Duncan ; Sleigh, Andy Sponsor: EPSRC ; Arup
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available