| Sažetak (engleski) | The turbulent flow field in rivers induces local scour around bridge piers, which affects thesafety of the bridge by undermining the foundations and altering the effects of the flow on thestructure. A typical scour protection measure for bridges built in large rivers is riprap slopingstructure, formed by a launchable stone conically shaped around the pier. Although the riprap sloping structure effectively protects the pier from scour, its considerable dimensions additionally obstruct the flow in the bridge opening and induce scouring downstream of the bridge profile. Shifting the inception point of erosion downstream of the pier means that the position and extent of a scour hole formed in the interaction of the oncoming flow and the coarse riprap are unknown. The deflected scour hole therefore poses an indirect threat not only to the bridge, but also to the infrastructure and river training structures downstream of the bridge as well. Numerous studies have been conducted to investigate the scour process around bridge piers in order to optimize bridge design. However, research on scour evolution around bridges protected with riprap throughout their lifespan, considering continuous changes in flow environment and bed development, are rare. Aiming to bridge the existing knowledge gap in scour development next to the piers protected with riprap sloping structure, this dissertation provides a detailed analysis of the local flow field and scour formation next to the riprap sloping structure under different flow scenarios. Scour depth is usually described as a function of various global flow parameters: mean velocity, mean depth, bed shear stress, etc. However, this approach neglects the local hydrodynamic processes resulting from the complex interaction between the irregularly shaped riprap and the turbulent nature of a river flow. The flow hydraulics of bridges that are protected by a riprapsloping structure is complex, characterized by a flow around the structure during water levels below the top of the riprap, and overtopping flow during high flow events when the structure is completely submerged. The main assumption behind this research is that the scour processes around the riprap sloping structure can be estimated more accurately if local flow parameters are used instead of the commonly used globally averaged values. The flow parameters defined as local in this dissertation are the ones measured in the near-bed surface layer above the scour hole and adjacent to the bridge pier above the riprap crest. The locations where the local parameters were measured are based on practical considerations of monitoring techniques in the field application. The aim of the dissertation is to compare the accuracy of scour hole depth estimation calculated with empirical equations using the relevant global and local parameters as input. The main assumption of the research is that scour hole depth is influenced by the local parameters, and that understanding of local flow environment can help to improve the accuracy of equilibrium scour depth estimation. To determine the most relevant parameters that influence scour next to the riprap sloping structure, the hydrodynamic parameters were measured on a physical model and simulated using the Flow-3D numerical model with identical geometry and flow environment. The physical model was set-up in a hydraulic channel with a sand bed and characteristic riprap sloping structure geometry. Clear-water scour experiments were conducted for a range of flow conditions representative for submerged and non-submerged conditions. The boundary conditions were adjusted to reflect low to mean water levels when the riprap sloping structure is not submerged, and high flow levels when the top of the riprap is submerged 2 times its height. A total of twelve experimental runs were carried out: seven clear-water cases (flow rates ranging from 20 l/s to 32 l/s), two marginal cases representing the no-scour and live-bed limits (18 l/s and 34 l/s), and three synthetical cases where flow velocity was kept constant (28.2 l/s, 11.1 l/s, and 9.1 l/s). Each experiment was carried out continuously for 30 hours to reach an equilibrium scour state, after which the bathymetry of the scoured bed was measured using an optical laser scanner. The instantaneous velocities in the near-bed surface layer above the scour hole were measured with the Acoustic Doppler Velocimeter Profiler (ADVP). A grid of 15 measurement points was defined, covering the extent of the area where scour hole was developing. Second order turbulence transport equations were calculated from the instantaneous velocity measurements to determine the value of the local Reynolds shear stress and turbulent kinetic energy. Based on the collected database combining the experimental and simulated data, regression models are used to evaluate direct relationships between the relevant hydrodynamic variables(flow depth, flow velocity, turbulent kinetic energy, and Reynolds shear stress) and scour depth. The coefficients of determination (R²) were used as a measure to quantify how well the regression model fits the data. The analysis was performed individually for the measurement point closest to the maximum scour depth and combined for multiple points spanning over the entire scour hole perimeter to determine whether the prediction of scour is more reliable when the hydrodynamic parameters are measured at a one point of maximum scour depth or averaged over the larger area. The coefficients of determination were calculated for all measurement points and for all experiments to determine the strongest dependence between scour depth and area size for the averaging the hydrodynamic parameters. The area that showed the strongest relationship with scour depth is the entire area within the scour perimeter, which consisted of 9 measurement points. The strongest individual relationship with scour hole depth is observed for the turbulent kinetic energy, followed by the Reynolds shear stress in the near-bed surface layer above the scour hole, 0.9436 and 0.9255, respectively. The lowest R² value was obtained for the local shear stress adjacent to the bridge pier above the riprap crest (0.5524). In general, the global parameters showed a weaker dependence on scour depth compared to the local parameters, as R² for mean flow depth, mean flow velocity and bed shear stress yielded lower values of 0.8964, 0.8855 and 0.8845, respectively. For the analysis of the empirical equations, the relevant parameters were categorized into three groups: global parameters averaged over the entire cross section, local parameters adjacent to the bridge pier above the riprap, and local parameters in the near-bed surface layer above the scour hole. The dimensional analysis is performed to ensure the dimensional homogeneity of the relevant parameters. Dimensional analysis is used to determine a functional dependency between scour as a response variable and the relevant parameters arranged in non-dimensional functional groups. As a result, three empirical equations were derived based on a multiple non- linear regression analysis. The performance of the derived equations was evaluated by plotting the predicted vs observed values against the line of best fit. The equation developed based on local parameters collected in the near-bed surface layer above the scour hole showed the best performance (R² = 0.9225), followed by the equation derived from parameters collected adjacent to the bridge pier above the riprap (R² = 0.9073), while the lowest performance was observed for the equation derived for global parameters (R² = 0.8760). Deriving an additional equation by modifying the best-fit equation to include the turbulent kinetic energy in the near- bed surface layer above the scour hole showed a significant improvement (R² = 0.9666) in predicting scour depth. When post-scour bed development is observed, its pattern shows a significant development of bedforms downstream of the scour hole for all clear-water experiments. The bedforms develop in the entire downstream area, beyond the direct influence of riprap on the flow field. This indicates that the scour hole, once it develops, initiates bed deformation that interacts with the flow field and influences further scour development. This phenomenon confirms the importance of considering local fluctuations as the main cause of local scour around riprap sloping structure. In this dissertation, two hypotheses were proposed to investigate the phenomenon of scour around bridge piers protected with a riprap sloping structure. The first hypothesis states that the equilibrium scour depth is influenced by the flow regime over the riprap structure. The first hypothesis is confirmed through the validation of the strong dependence of the local parameters above the riprap structure adjacent to the bridge pier (R² = 0.9073) compared to the other empirical equations. The second hypothesis states that the maximum scour depth does not correspond to the maximum flow depth. The second hypothesis is confirmed through the validation, in which the scour hole reached a larger extent during low flows, where the overflowing effect above the riprap is absent, compared to the experiments at high flow, when the water level is above the riprap, while at the same time the mean flow approach velocity was kept constant. This research identified the most relevant hydrodynamic parameters needed to accurately estimate the scour hole depth developing next to the riprap sloping structure and thus improving the understanding of the scour process. The empirical equations demonstrate a significant improvement in the prediction of downstream morphodynamic processes when local parameters are considered, compared to the traditional approach that considers global parameters. The dominant influence of the local flow field above the scour hole on the morphodynamic process emphasizes the importance of monitoring local parameters in real time to support the decision-making process in bridge management systems. |
