Doctoral Degrees (Civil Engineering)

Permanent URI for this collection

https://scholar.sun.ac.za/handle/10019.1/764

Browse

Browsing Doctoral Degrees (Civil Engineering) by browse.metadata.advisor "Bosman, Adèle"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Determination of the critical incipient failure conditions for angular riprap dumped on wide & steep trapezoidal channels
    (Stellenbosch : Stellenbosch University, 2019-04) Appolus, Michael; Bosman, Adèle ; Basson, G. R.; Stellenbosch University. Faculty of Engineering. Dept. of Civil Engineering.
    ENGLISH ABSTRACT: The main objective of this thesis was to determine the critical MN that defines the incipient failure conditions of angular riprap dumped on wide and steep trapezoidal channels. A total of 32 physical hydraulic model tests were performed in three test series. There were 7 tests performed for Test series one, 15 tests performed on Test series two and 10 tests were performed on Test series three. The tests were executed by gradually increasing flow rates over the hydraulic model to enable establishment and recording of the flow rate that induced incipience of riprap for a specific hydraulic model setup. Failure was defined as the flow rate that instigated a significant movement of riprap stones less and equal to D50. Based on the physical model tests of this thesis it was found that for the riprap on the bed of a relatively wide trapezoidal channel (bottom width to D50 ratio of 16 to 31) and steep bed slopes (of 0.333-0.5), the critical MN value defining the incipient failure conditions for these steep bed slopes was 0.12 with an exceedance probability of 95%. This MN value is in good agreement with Rooseboom’s (1992) MN criteria of 0.12. In addition, the MN for defining the critical incipient failure condition of riprap on a 0.4 steep side bank slope was found to be 0.227, with an exceedance probability of 95%. Based on the HEC-RAS steady state flow numerical simulations of the physical model tests series performed in this thesis, it was found that HEC-RAS overestimates the actual incipient failure MN. HEC-RAS overestimated the critical incipient failure MN of the steep bed and steep side bank by a critical factor of 1.91 and 1.35, respectively. As a result, the two factors were recommended as the MN adjustment factors (the steep bed and side bank MN must be adjusted to MN values of 0.12 and 0.227, respectively) for defining the incipient failure of a specific D50 rock size when using HEC-RAS steady state flow analysis. Lastly, the applicability of the findings of this study are limited to riprap dumped in straight trapezoidal cross-sectional channels with steep beds ranging from 0.333 to 0.5 and with side bank slopes of 0.4. The scale of the hydraulic physical model used in the investigation was selected relatively large i.e. 1:15 to minimize model scale effects. The model D50 size was 0.038 m and 0.075 m which represent prototype stone sizes with D50 between 0.57 m and 1.125 m respectively. The results of the study are therefore only valid for the design of prototype D50 stone size between stone 0.57 m and 1.125 m. Most importantly, the bed bottom width to D50 ratio needs to be between 16-31.
  • Thumbnail Image
    Item
    Investigation of novel deflector shapes for uncontrolled spillways
    (Stellenbosch : Stellenbosch University, 2024-02) Wright, Henry John; Bosman, Adèle ; Brink, Isobel ; Stellenbosch University. Faculty of Engineering. Dept. of Civil Engineering.
    ENGLISH ABSTRACT: The hydraulics of stepped spillways are generally well understood, although numerous fundamental hydraulic aspects remain inadequately explored. Critical knowledge gaps persist, including aerated flow hydraulics, hydraulics of embankment flows, hydraulics of stepped spillways for steep gravity dams, environmental hydraulics as well as turbulent interactions between cavity flow and skimming flow. Notably, the elusive safe unit discharge limits for stepped spillways remain undefined, with conflicting findings in the literature. The majority of stepped spillways have been designed for a maximum unit discharge of 25 to 30m3/s/m due to the risk of cavitation damage. It has further been reported that the critical velocity of approximately 20 m/s for the inception of cavitation on stepped spillways is obtained at a unit discharge of 25 m3/s/m. Further research in the field revealed that a bottom aerator becomes imperative for discharges greater than 30m3/s/m. However, discrepancies persist, with other researchers suggesting that the safe unit discharge is lower, quoting 11.5 m3/s/m to 14 m3/s/m for step heights of 0.6 m to 1.2m, respectively. Therefore, the exact limits of stepped spillways remain unquantified when water flows on the downstream slope. In China, the Flared Gate Pier (FGP) has been used on stepped spillways, particularly the X-type and Y-type piers. These piers support the crest gates and have been customised to contract the flow rapidly into a high-velocity jet. These piers have been used at, amongst others, the Dachaoshan Dam, a 111 m high Roller Compacted Concrete (RCC) gravity dam, with a maximum unit discharge designed of 193m3/s/m. These piers redirect flow into high-velocity jets, achieving efficient energy dissipation without relying on the stepped spillway face. Although historically utilised exclusively with gated spillways, FGPs hold potential as deflector-type energy dissipaters and were used as the basis for the novel deflector investigations in this research. To date, a variety of aerators have been fitted to improve spillway performance. Other aeration methods, such as the use of Roberts splitters, rectangular protrusions and triangular protrusions have been proposed, with some of these designs being successfully implemented. However, research has noted that these methods yield only marginal increases in the safe unit discharge of stepped spillways. The main concern regarding stepped spillways is the cavitation risk during high discharges, with a critical cavitation parameter of 0.5 compared to 0.2 for smooth chutes. This limits the maximum allowable unit discharge. While cavitation pitting has not been reported on prototype spillways, the exact conditions under which cavitation on stepped spillways may occur remain uncertain. The current research investigated the feasibility of a novel deflector form aimed at increasing the safe discharge capacity of spillways by deflecting the flow away from the spillway slope. The research incorporated a comprehensive approach, comprising a series of numerical models to simulate the hydrodynamic environment as well as four physical models. Numerical model simulations were undertaken with FLOW-3D HYDRO® and ANSYS FLUENT® computational fluid dynamics (CFD) software to optimise the deflector geometries before being tested with a physical model. A 1:50 scale physical model was constructed to investigate the influence of different deflector shapes. The investigation spans a range of prototype unit discharges ranging from 50 to 200m3/s/m and evaluates factors such as water surface profiles created by the deflector and pressure distribution on the deflector. A regression analysis was performed on the collected physical model data to develop equations that predict the jet's inner and outer trajectory and jet breakup length. The proposed novel deflectors developed in this study proved to be effective at various flow rates when the flow trajectory and threshold pressures were considered. These deflectors could be used for dams higher than 150 m and unit discharges ranging between 100 and 200 m3/s/m. Further research is required to improve, amongst others, deflector geometries, to study variables and to undertake additional measurements to conform and improve the efficiency of the novel deflectors, using this research as a basis.
  • Thumbnail Image
    Item
    Optimisation of sand trap and settler designs for efficient deposition of suspended sediment
    (Stellenbosch : Stellenbosch University, 2024-03) Mc Leod, Claudia; Bosman, Adèle; Smit, G.J.F; Stellenbosch University. Faculty of Engineering. Dept. of Civil Engineering.
    ENGLISH ABSTRACT: Sand traps and settlers are used at river abstraction or diversion works for the control of sediment loads for potable, irrigation and hydropower usage. The performance of a sand trap and settler is judged by its capability to sufficiently deposit suspended sediment particles and its flushing capability of bed load sediment. The design properties of a canal type sediment trap, such as the depth of flow, total length, cross-section, slope and inlet position, mainly determine its hydraulic efficiency. One of the main objectives of this study was to develop hydraulic design guidelines for sand traps and settlers for river conditions that have high loads of fine non-cohesive sand. This study investigates design considerations (dimensions, slope, cross-section, type of intake locations, and sediment intake concentration as well as inlet designs) of sediment traps by using an existing fully three-dimensional Computational Fluid Dynamics (CFD) model coupled in terms of hydrodynamics and sediment transport developed by Sawadogo (2015). The hydraulic performance of an existing sand trap and settler in Southern Africa was also investigated to identify hydraulic design aspects that could be improved. This was achieved by performing field measurements, analysing the field results of sediment deposition and velocities within the traps, and numerically investigating the case studies to recommend possible design upgrades to improve the efficiency of the traps. This study also investigates the innovative “Split-and-Settle” sand trap concept, initially proposed by Støle in 1993, by means of a three-dimensional CFD model as a second main objective. The “split-and-settle” approach directly refers to dividing the flow in a sand trap into sediment-free and sediment-laden water and then removing the sediment from the water. As sediment-laden water traverses a canal, the suspended sediment concentration increases near the bottom along the length of the canal as sediment tends to deposit. The split-and-settle concept leverages this sediment concentration gradient by dividing the flow into upper and lower parts. The concept was investigated by conducting physical modelling to generate data in a controlled environment to calibrate a numerical model. To calibrate and evaluate the sensitivity of the numerical model, appropriate parameters were adjusted until a good agreement was reached between the physical and numerical model results for the suspended sediment concentration. The parameters included the convection-diffusion equation’s turbulent Schmidt number, mesh configuration for capturing the split plate boundary and turbulent intensities. The principal contribution from this work is the calibrated and validated numerical model which could be further used to propose design guidelines for the split-and-settle sand trap. Additional benefits of the Split-and-Settle sand trap design include that it reduces the required length of sand traps and therefore are more economical whilst being hydraulically efficient, sustained operational capacity for handling substantial volumes of sediment-laden water, and minimal maintenance demand due to continuous flushing. Moreover, it operates without the inherent risk of scour holes that can become blocked or clogged and reduce efficiency. Ultimately, this research underscores the potential of the Split-and-Settle sand trap as a valuable tool in sediment management and hydraulic engineering.