Laboratory Investigation of Mean Drag in a Random Array of Rigid, Emergent Cylinders

• Published in: Journal of hydraulic engineering (New York, N.Y.) Vol. 134; no. 1; pp. 34 - 41
• Main Authors: Tanino, Yukie, Nepf, Heidi M
• Format: Journal Article
• Language: English
• Published: Reston, VA American Society of Civil Engineers 01.01.2008
• Subjects: Applied sciences; Buildings. Public works; Exact sciences and technology; Hydraulic constructions; TECHNICAL PAPERS; Experimental data; Two-dimensional flow; Open channel flow; Experimental method; Aquatic plants; Theoretical study; Two dimensional flow; Experimental study; Experimental result; Drag; Vegetation; Laboratory test; Aquatic plant; Cylinders
• ISSN: 0733-9429; 1943-7900
• DOI: 10.1061/(ASCE)0733-9429(2008)134:1(34)

This paper investigates the drag exerted by randomly distributed, rigid, emergent circular cylinders of uniform diameter d . Laboratory measurements are presented for solid volume fraction ϕ=0.091 , 0.15, 0.20, 0.27, and 0.35 and cylinder Reynolds number Rep ≡ Up d∕ν=25 to 685, where Up =temporally and cross-sectionally averaged pore velocity and ν =kinematic viscosity. These ranges coincide with conditions in aquatic plant canopies. The temporally and cross-sectionally averaged drag coefficient, CD , decreased with increasing Rep and increased with increasing ϕ under the flow conditions investigated. The dimensionless ratio of the mean drag per unit cylinder length ⟨ fD ¯ ⟩H to the product of the viscosity, μ , and Up exhibits a linear Rep dependence of the form ⟨ fD ¯ ⟩H ∕(μ Up )= α0 + α1 Rep , consistent with Ergun’s formulation for packed columns. In the range of experimental conditions, α1 increases monotonically with ϕ . In contrast, α0 is constant within uncertainty for 0.15⩽ϕ⩽0.35 , which suggests that viscous drag per unit cylinder length is independent of ϕ in this range.