Hybrid analysis of river training structures: integration of 3D numerical modeling with field-scale hydrodynamic assessment
摘要
In this paper, the performance of groynes and levees as river training structures on the Aras River in Ashtabin village is examined using three numerical models: FLOW-3D, HEC-RAS 2D and CCHE2D. The study was based on a 3.5 km stretch whereby the flow velocities varied within the range of 0.2 to 2.5 m/s with a range of depths of 0.2 to 5 m. It was found that groynes raised the local flow velocities by 45% at the heads, but that levees raised the main channel velocity by 2535%. The measurements of turbulent kinetic energy demonstrated a higher turbulence of up to 80% increase in groynes than natural conditions and an increase of 30% in levees, which is moderate. Peak scour depths of groyne heads were 2.5-3 times groyne length, with levees having a less homogeneous, shallower scour, in the range of 1.5 times the depth of flow. The numerical modeling of the results revealed that FLOW-3D gave the best results that had less than 5% error in predicting the surface and bed shear stresses when compared with HEC-RAS 2D and the CCHE2D that had 15% and 25% errors, respectively. The experiment also found out that groynes formed powerful horseshoe vortices of relative intensities 0.4–0.6 of mean flow velocity in comparison to the weak secondary flows (0.2–0.3) around levees. The results are quantitative information useful in designing the best river training structure and the advantages of three-dimensional modeling in complex hydraulic modeling.