Author(s): Hanwen Cui; Stefan Felder; Matthias Kramer

Linked Author(s): Hanwen Cui, Stefan Felder, Matthias Kramer

Keywords: Rtificial grass; Flow resistance; Hydraulic design; Velocity measurement; Vegetated spillways

Abstract: Grass linings on spillways and waterways with mild to moderate slopes can provide ecological advantages such as groundwater replenishment and vegetation growth, while maintaining the functionality of flow conveyance. Design of chutes with vegetation and grass linings have mostly been based upon soil erosion considerations, and simplified flow resistance models that did not consider free-surface aeration. A critical parameter for the hydraulic design of grass-lined chutes is the estimation of flow resistance. The flow depth and the depth-averaged velocity are two key parameters that need to be considered in the assessment of channel stability and flow resistance. The flow velocities over grass-lined channels are directly influenced by the grass canopy drag, and a consideration of the detailed velocity profiles is needed to compute the depth-averaged flow velocity. Previously, a four-layered semi-analytical velocity superposition model has been proposed for subcritical flow condition, whereas limited information exists on the velocity distributions of aerated supercritical flows. To address this research gap, the current study re-analyzed a multi-phase flow data set collected in a laboratory chute of 10.8 degrees with artificial grass lining, consisting of a wide range of discharges between 0.031 m2/s and 0.313 m2/s. An additional layer of constant interfacial velocity at the air-water interface expanded the four-layered subcritical velocity profile model. Considering the resulting five-layer superposition model, an integration of the individual velocity components provided an estimate of the depth-averaged velocity. The contribution of each of the five velocity layers on the depth-averaged velocity was tested, for potential simplifications of the integration. This analysis identified dominant velocities in the mixing layer and the log law layer, while uniform layer, wake function and free-surface layer had smaller contributions. Through comparison, simplifications were suggested to the integration limit of the log-law layer and the wake function, leading to shortened explicit equations for mean velocity and friction factor. Overall, the present study shed novel insights into the prediction of the mean velocity and friction factor, which may assist practitioners in the design of grassed channels without conducting mass field tests.

DOI: https://doi.org/10.3850/IAHR-39WC252171192022581

Year: 2022