A RANS numerical study of experimental swash flows and its bed shear stress estimation

• Published in: Applied ocean research Vol. 100; pp. 102145 - 12
• Main Authors: Hu, By Peng, Xie, Jiafeng, Li, Wei, He, Zhiguo, Marsooli, Reza, Wu, Weiming
• Format: Journal Article
• Language: English
• Published: Barking Elsevier Ltd 01.07.2020; Elsevier BV
• Subjects: Backwash; Bed shear stress; Bottom stress; Boundary layers; Dam failure; Depth; Flow structures; Hydrodynamics; RANS modeling; Shallow water; Shear stress; Spatial variations; Surf zone; Swash flow; Vertical flow; Vertical mixing; Vorticity; Water depth; Swash flow; RANS modeling; Bed shear stress
• ISSN: 0141-1187; 1879-1549
• DOI: 10.1016/j.apor.2020.102145

Quantification of the bed shear stress in the swash zone is subject to much uncertainty due to the very shallow water, complex flow structure and the strong unsteadiness in this zone. This study presents a RANS modeling of experimental dam-break flow generated swash processes, giving numerical reproduction of vertical flow structure of swash flows and thus an opportunity to study the magnitude and features of the bed shear stress. Firstly, numerical solutions (i.e., temporal or spatial variation of the depth-averaged velocity and water surface elevation, the time variations of the swash front, the vorticity structure and the boundary layer thickness) are compared with existing laboratory data and the previous knowledges on swash hydrodynamics. Secondly, using the computed vertical flow structure as input data, performances of four methods (the Reynolds stress method, and the three variants of the log-law method: the depth-averaged variant, the one-position variant, the two-position variant) for estimating the bed shear stress are evaluated. It is shown that the Reynolds stress method produces non-zero bed shear stress during the reversal time, during which the swash flows may be partly stagnant. Among the three variants of the log-law method, the depth-averaged method appears to agree with the measured data much better than the other two variants: those by the two-point log-law method is very sensitive to the selected near-bed positions, and those by the empirical value variant may underestimate (overestimate) the bed shear stress in the early uprush stage (in the initial stage of the backwash).