Numerical simulation of scouring around groups of six cylinders with different flow directions

ABSTRACT

Scouring around the pile foundation is one of the potential failures of offshore bridges that must be assessed due to the possibility of damaging the bridge structure. However, there is still few in-depth research regarding local scour simulations around groups of six cylinders consisting of more than one row, which is most widely used for group arrangement performed numerically. The maximum scour depth and its mechanism surrounding the six-cylinder group were determined by numerical simulation of scour consisting of multiple rows arranged side-by-side and in tandem with different pile spacing ratios using Flow-3D software. A validation study using experimental investigation was completed to verify the accuracy of the computational model. Numerical simulations employing the Van Rijn transport rate Equation and the RNG k – ɛ turbulence model produced results for scour depth evolution and bed elevation contour, which were in accordance with the experimental study. The numerical results show that the best arrangement for the minimum scour depth of the six-cylinder group is the tandem arrangement of two rows and three columns with pile spacing ratio of 3.5. As a result of this study, engineers can optimize the design of bridge pile foundations to enhance safety and economic efficiency.

CO EDITOR-IN-CHIEF:

• Chen, Jeng-Tzong

ASSOCIATE EDITOR:

• Ou, Yu-Chen

KEYWORDS:

• Numerical simulation
• local scour
• pile group
• Flow-3D

Disclosure statement

No potential conflict of interest was reported by the author(s).

Nomenclature

$A_i$	=	Area fraction
$c_{b,i}$	=	Volume fraction of species i
$CDIS1$	=	Production coefficient
$CDIS2$	=	Decay coefficient
$CDIS3$	=	Buoyancy coefficient
Crough	=	Wall roughness
D	=	Pile diameter
$d_i$	=	Sediment grain size
$Dif{f}_{k_T}$	=	Diffusion term of turbulent kinetic energy
$Dif{f}_{\epsilon }$	=	Diffusion term of dissipation
$d_{\ast ,i}$	=	Sediment grain size (dimensionless)
d50	=	Average grain size
$f_i$	=	Viscous acceleration
$g$	=	Absolute value of gravity
G	=	Pile spacing
$G_T$	=	Turbulent buoyancy
$G_i$	=	Body acceleration
H	=	Height of channel
ks	=	Nikuradse roughness
$k_T$	=	Kinetic energy
k – ɛ	=	K-epsilon turbulence model
L	=	Length of channel
Lh	=	Entrance length
$p$	=	Average hydrodynamic pressure
$P_T$	=	Production of kinetic energy
$s_i$	=	Ratio of ρi and ρf
$t$	=	Time
$uvw$	=	Fluid velocity section in stream-wise, transverse and vertical directions
$V_F$	=	Volume fraction
$v_f$	=	Fluid kinematic viscosity
W	=	Width of channel
x-, y-, and z-	=	Coordinates in stream-wise, transverse and vertical directions
${\beta }_{VR,i}$	=	Bed-load coefficient
${\epsilon }_T$	=	Rate of turbulent energy dissipation
${\theta }_{cr,i}$	=	Critical shields number
$\rho$	=	Fluid density
${\rho }_i$	=	Sediment grain mass density
${\tau }_{cr,i}$	=	Critical bed shear stress
${\mathrm{\Phi }}_i$	=	Bed-load transportο
$\mathrm{\partial }$	=	Partial derivatives