Numerical Simulation of Regular Waves onto a Vertical Circular Cylinder

Zhenghao Liu*, Decheng Wan‡

State Key Laboratory of Ocean Engineering, School of Naval Architecture, Ocean and Civil Engineering,
Shanghai Jiao Tong University, Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration,
Shanghai 200240, China
*
Presenting author: [email protected]
‡
Corresponding author: [email protected]
http://dcwan.sjtu.edu.cn

Abstract
Wave run-up phenomenon is of great significance during the design of a fixed or floating
structure. In this work, the wave run-up on a truncated surface-piercing circular cylinder is
investigated using a numerical wave tank. The numerical simulations are carried out by the in-
house CFD solver naoe-FOAM-SJTU which is developed on the open source platform
OpenFOAM. The volume of fluid (VOF) method is applied to capture the free surface. The
surface elevation around the cylinder is probed by a series of wave gauges and analyzed using
the Fourier analysis. The response amplitude operates (RAOs) of surface elevation are
presented and compared with experimental data. Reasonable agreement shows the present
solver is capable to investigate the wave run-up on a cylinder. The local surface elevation
around the cylinder, the wave force and the scattered wave field around the cylinder is also
investigated in detail.
Keywords: wave run-up; truncated vertical cylinder; naoe-FOAM-SJTU solver; surface
elevation

Introduction
Offshore structures such as Spars, semi-submersibles and tension leg platforms (TLP) have
been widely applied in ocean engineering. Wave run-up on the columns of the structures can
be relatively large and even causes green water on deck in severe environment. Generally,
wave run-up height is defined as the maximum vertical wave elevation to the still water
surface. Significant nonlinear wave-structure interactions can be observed during the wave
run-up process. The accurate prediction of wave run-up is of great importance for the air-gap
design of offshore structures. Numerous researchers have done experimental, theoretical and
numerical studies on wave run-up onto piles, sloped beaches and columns of both fixed and
floating offshore structures.

Various experimental investigations of wave run-up on vertical cylinders have previously

been performed. Galvin and Hallermeier (1972) experimentally studied the wave run-up on a
cylindrical column for the first time [1]. A series of wave gauges were mounted near the
column to obtain the distribution of free surface around the column. It was found that when
the waves pass through the vertical column, two important factors will affect the wave run-up
effect: 1) scattering effect due to wave-structure interaction; 2) viscous dissipation effect at
the column wake region. Chakrabarti and Tam (1975) conducted a series of model tests to
investigate the regular waves onto a large-scale cylinder [2]. The incident wave steepness kA
is 0.03~0.19 and the scattering parameter kr is 0.34~1.55, respectively. Their work was
focused on the wave run-up phenomenon and the effect of incident wave angle on the
distribution of the cylinder surface pressure, while the free surface around the cylinder was
not discussed. Morris-Thomas et al. (2002) investigated the effect of wave steepness and the
scattering parameter on the wave run-up on a fixed cylinder in a towing tank [3]. Nielsen
(2003) firstly studied the effect of cross-sectional shape on the wave run-up and free surface
of cylinders [4]. Compared with circular cylinder, stronger nonlinear interaction can be found
for circular-like cylinder. Systematic model tests were conducted in Shanghai Jiaotong
University to investigate the effect of aspect ratio, cross-sectional shape, wave parameters and
current on wave run-up phenomenon in 2010.

Besides experimental research on wave run-up, the nonlinear wave-structure interaction

phenomenon has been studied by numerous researchers through theoretical methods.
Generally, the theoretical methods are based on potential theory in which an idealized fluid
domain is assumed. The Laplace equation is solved with applied boundary conditions to yield
a velocity potential. The free-surface elevation around the column can be obtained by
application of the unsteady Bernoulli equation and velocity potential at the free-surface
position. In the early days, the approximate results of wave run-up on a single cylinder
according to first- and second-order potential flow theory [5][6]. Trulsen and Teigen (2002)
developed fully nonlinear numerical wave tanks (NWT) to investigate nonlinear interaction
between wave and a truncated cylinder, and compared the predicted results with experimental
data [7]. They found the effects of viscosity may take account for the discrepancies between
the theoretical method and experiment. Similar researches can be found in [8]. Morris-
Thomas and Thiagarajan (2004) adopted the linear diffraction theory and the commercial
software WAMIT to predict the wave run-up around a cylinder [9]. They suggested that linear
diffraction theory is insufficient for wave run-up estimation. This confirms conclusions made
by previous authors. The second-order harmonic components predicted by WAMIT show
reasonable estimates of wave run-up when the scattering parameter is small.

With the rapid development of computer technology during the last several decades,
computational fluid dynamics (CFD) method has been widely applied in ocean engineering
field. Previous numerical simulations based on CFD have also been performed on wave run-
up problems. Based on the open source platform OpenFOAM, Cao and Wan (2014, 2015,
2017) simulated the regular and solitary waves on to a circular cylinder, and the comparison
of numerical results of wave run-up and experimental data showed the reasonable agreement
[10]-[12]. Donald G Danmeier et al. [13] used the ComFLOW software to simulate the wave
run-up around a semi-submersible platform. Sun et al. [14] used a both potential flow solver
DIFFRACT and a full CFD solver OpenFOAM to investigate nonlinear interactions between
regular waves and a single truncated circular column. The predicted free surface elevation
around the column and the wave forces were analyzed and compared with experimental data.

To investigate wave run-up, the ITTC committee organized several studies, including
experimental and numerical researches in 2013. The results of a series of model tests for a
truncated circular column in regular waves at MARINTEK and MOERI are used in the ITTC
benchmark study [15]. The time histories of surface elevations and wave forces were provided.

The objective of the present work is to investigate the wave run-up on a surface-piercing
circular cylinder. The computations in this paper are performed with the in-house CFD solver
naoe-FOAM-SJTU, which is developed based on the open source code OpenFOAM. The
numerical results of the free surface elevation around the column and the wave forces are
presented and compared with experiments performed at MOERI. The details information of
wave run-up around cylinder will be given and discussed. The results show that the present
approach can be an alternative tool to deal with nonlinear wave-structure interactions.
Numerical methods
naoe-FOAM-SJTU solver
The CFD solver naoe-FOAM-SJTU is designed for computing viscous flows around ships
and ocean structures [16-20] and mainly composed of a dynamic deforming mesh module, a
6DoF motion module, a velocity inlet wave-making module and a mooring system module.

In present study, the incompressible unsteady Reynolds averaged Navier-Stokes (URANS)

equations are adopted as the governing equations, and can be written as follows:

∇ ⋅U = 0 (1)
∂ρU
+ ∇ ⋅ ( ρ (U − U ɡ )U ) = −∇pd − g ⋅ x∇ρ + ∇ ⋅ ( µeff ∇U ) + (∇U ) ⋅∇µeff + fσ + f s (2)
∂t

where U and Ug are the velocity field and the velocity of grid nodes, respectively. pd is the
dynamic pressure and p is the total pressure, ρ is the mixed density of the two phases water
and air. μeff is the effective dynamic viscosity, in which 𝜈𝜈 and 𝜈𝜈t are kinematic viscosity and
eddy viscosity, respectively. fσ is the surface tension, which impacts the free surface. fs is a
source term, added to generate the sponge layer for wave absorbing.
Capture of Free Surface
The free surface of the two-phase flow is captured by the volume of fluid (VOF) method [21]
with artificial bounded compression techniques. The method is based on a volume fraction α
which can control numerical diffusion and capture the interface with high resolution. The
volume fraction function can be determined by solving a transport equation:

∂α
+ ∇ ⋅ [(U − U ɡ )α ] + ∇ ⋅ [U r (1 − α )α ] = 0 (3)
∂t

The first two terms on the left-hand side of Eq. (3) stand for traditional volume of fluid
transport equation while the third term represents the artificial compression term. The velocity
field Ur compressing the interface is computed at cell faces by the maximum velocity
magnitude at the interface region:

{
U r , f = n f min Ca
f
Sf
, max ( )}
f
Sf
(4)

where φ is face volume flux; Cα is a compression coefficient controlling the magnitude of

compression, in this paper it is chosen to be 1.0. Larger value will increase the compression of
the interface, leading to larger detrimental velocity gradients around the interface. The
compression term only works on the interface without affecting the numerical computation
out of the transition layer due to term (1−α)α.

Different phases are marked using volume fraction α which indicates the relative proportion
of fluid in each cell, it is defined as Eq. (5). For an interface cell, the value of volume fraction
α is between 0 and 1, representing it contains both water and air.
 a =0 air

a =1 water (5)
0 < a < 1
 interface

In physical domain, the density of fluid ρ and the dynamic viscosity µ can be obtained by a
weighted value based on the volume fraction α:

ρ= αρ1 + (1 − α ) ρ 2 (6)
µ= αµ1 + (1 − α ) µ2 (7)

Where ρ1 and ρ2denote the density of water and air, μ1 and μ2 denote the viscosity coefficient
of water and air, respectively.
Wave Generation and Damping
Our naoe-FOAM-SJTU solver includes a wave generation and damping module. To generate
the Stokes first deep water wave, the inlet boundary conditions are set as follows:

=η a cos(k ⋅ x − ωet + δ ) (8)

u = U 0 + aω e kz cos(k ⋅ x − ωt + δ ) (9)
=v aω e kz cosβ cos(k ⋅ x − ωt + δ ) (10)
=w aw e kz sinβ sin(k ⋅ x − wt + δ ) (11)

Where ζ is transient wave elevation, a, ω and k are wave amplitude, wave frequency and
wave number, respectively. U0 is hull speed.

To avoid wave reflection, a sponge layer is setup at the outlet of the computational domain.
The term of fs is introduced into Eq. (2) for absorption of waves and defined as:

 x - xs 2
- rα（ ）(U - U ref )
f s ( x) = 
s
Ls (12)
0


Where in which, ρ is the water density. The αs is an artificial viscosity coefficient controlling
the intensity of the sponge layer. The xs is the coordinate of the start position of the sponge
layer and Ls is the length of the sponge layer. The source term fs has no effects on the domain
out of the sponge layer.
Discretization schemes
The RANS and VOF transport equations are discretized by finite volume method (FVM). Van
Leer scheme is applied for VOF equation in OpenFOAM. The PIMPLE (merged PISO-
SIMPLE) algorithm is used to solve the coupled equation of velocity and pressure. The
convection terms are solved by a second-order TVD limited linear scheme, and the diffusion
terms are approximated by a second-order central difference scheme.

Computational model and test conditions

Computational model
To simulate wave run-up on a truncated surface-piercing column, the cylinder model at full
scale from MOERI is selected in numerical simulation. The radius of the cylinder is R = 8.0 m,
and the draft is 24.0 m. A series of wave probes in both experiment and numerical simulation
are shown in Figure 1, and the corresponding location are given in Table 1. The distances
between inner circle and outer circle are 0.2063 m and 8 m, respectively.

Figure 1 Layout of wave probes

Table 1 Location of wave probes

Inner circle x (m) y (m) Outer circle x (m) y (m)
WPB1 -8.2063 0.0000 WPO1 -16.0000 0.0000
WPB2 -5.8027 -5.8027 WPO2 -11.3137 -11.3137
WPB3 0.0000 -8.2063 WPO3 0.0000 16.0000
WPB4 5.8027 -5.8027 WPO4 11.3137 11.3137
WPB5 8.2063 0.0000 WPO5 16.0000 0.0000

Figure 2 shows the arrangement of computational domain. The domain extends to -2L < x <
3L, -L < y < L, -L < z < 0.5L. Where L represents the wave length. The water depth is set as L.
The length of sponge length is L, starting from x = 2L. The vertical cylinder is fixed in the
center of the wave tank.

Figure 2 Computational domain

The computational mesh is shown in Figure 3. About 70 grids per wavelength and 20 per
wave height are applied. The total grid number is about 1.7 million. To make it easy to
converge in each time step, the interface Courant number was controlled to be under 0.3. The
time step is 0.001s in each case. The boundary conditions are as follows: Velocity inlet is
adopted. Zero-Gradient condition is applied at the outlet. The no-slip boundary condition is
imposed on the cylinder. The symmetry boundary condition is applied to the side walls.

Figure 3 Mesh of computational domain

Test conditions
The incident wave conditions were set up according to the benchmark study conditions
proposed by the 27th ITTC committee. The Stokes first order deep water wave is applied in
the present work. Two wave periods (T = 7s, 9s) have been investigated. Three steepness
parameters (H/L = 1/30, 1/16 and 1/10) were simulated for each wave period, where H is the
wave height and L is the wave length. Another parameter in our analysis is the scattering
parameter k0r, where k0 (k0 = 2π/L) is the wave number. Details of the wave conditions are
listed in Table2.

Table 2 Wave conditions

T=7s T=9s
H/L L(m) D/L H (m) L(m) D/L H (m)
1/30 2.548 4.212
1/16 76.44 0.21 4.777 126.36 0.13 7.898
1/10 7.644 12.636

Results and discussion

RAOs of surface elevations
The response amplitude operates (RAOs) are effectively transfer functions used to determine
the effect of wave on the ocean structures. The obtained time histories of free surface
elevation are generally analyzed to acquire the RAOs using the Fourier analysis. The acquired
RAOs of free surface elevations at ten wave probe locations from our CFD simulation are
compared with the experimental data from MOERI. The RAOs of free surface elevations
from potential flow solver DIFFRACT [14] are also adopted in this work. Figure 4 shows the
comparisons of the RAOs of surface elevation near the cylinder at WPB3 and WPB4 for the
condition wave period T = 7s and H/L =1/30, 1/16, 1/10. As shown in Figure 4(a), for small
wave steepness condition, both the CFD results and DIFFRACT results agree well with the
experimental data at WPB3. As the steepness increases, the CFD solver can give more
accurate prediction than the potential flow theory. This is more obvious for WPB4 in Figure
4(b), for the H/L = 1/10 condition, the CFD solver can give a much better agreement with the
experimental data. This may be the strong nonlinear interactions at the downstream location
WPB4. Reasonable agreement between the CFD results and experimental data implies that the
present solver is capable to predict the wave run-up on a truncated cylinder.
 (a)WPB3 (b)WPB4
Figure 4 Comparisons of RAOs of surface elevations for T = 7s

Time histories of wave probes

For the wave period T = 7s condition, time histories of the free surface elevation obtained
from our CFD simulation are shown in Figure 5 (inner circle probes) and Figure 6 (outer
circle probes). As can be seen from Figure 5, the wave probe in front of the cylinder reaches
the largest wave amplitude in any wave height condition. This may due to the wave-structure
interaction in front of the cylinder. When the wave crest reaches the cylinder, the water runs
up vertically along the cylinder. The surface elevation at WPB1 is about two times than the
incident wave amplitude. The surface elevation at the downstream quarter point WPB4 is the
smallest, and secondary crests can be found for steeper waves (H/L = 16 and 1/10). This may
imply strong nonlinear interaction at this location. After passing the sides of the column, the
incident wave encounters and interferes with each other, resulting free wave flows backwards
in the upstream direction and overlaps with the incident wave. Thus, the secondary crest can
be observed. Similar trends can be found for outer circle wave probes, as shown in Figure 6.
However, the discrepancies for these wave probes is reduced compared with the inner circle
ones. The secondary crests at WPO4 is not so obvious as WPB4 near the cylinder.

For the wave period T = 9s condition, time histories of the free surface elevation are shown in
Figure 7 (inner circle probes) and Figure 8 (outer circle probes). The missing of surface
elevation data at WPO4 for H/L = 1/16 condition is caused by inappropriate wave probe setup
in numerical simulation. Secondary wave crest can also be found at WPB4 in steep waves
(H/L = 16 and 1/10).

(a)WPB1 (b)WPB4 (c)WPB5

Figure 5 Time series of surface elevation for inner circle wave probes for T = 7s
 (a)WPO1 (b)WPO4 (c)WPO5
Figure 6 Time series of surface elevation for outer circle wave probes for T = 7s

(a)WPB1 (b)WPB4 (c)WPB5

Figure 7 Time series of surface elevation for inner circle wave probes for T = 9s

(a)WPO1 (b)WPO4 (c)WPO5

Figure 8 Time series of surface elevation for outer circle wave probes for T = 9s

Time histories of wave forces

Figure 9 and Figure 10 show the time histories of horizonal wave force for T = 7s and T = 9s,
respectively. The horizonal wave force increases with the wave height. The nonlinearity in
wave force is not as strong as surface elevation. This is because the local nonlinear effects are
integrated out when computing the force, while the surface elevation show the original local
nonlinearity of wave run-up the cylinder. So it is essential to study the local free surface near
the cylinder.

Figure 9 Time series of horizonal wave force for T = 7s

 Figure 10 Time series of horizonal wave force for T = 9s
Scattered wave field around the cylinder
Figure 11 shows the local free surface around the cylinder for wave at T = 7s and H/L =1/10.
The wave diffraction field around the cylinder can be clearly seen. When the wave crest
approaches the cylinder, the water is blocked in front of the cylinder and concentric wave
field (type1) can be observed. This is specified by Swan and Sheikh (2015) [22]. Then the
water bypasses the side of the cylinder, the non-concentric wave field (type2) is developed at
the downstream shoulders. This may induce strong nonlinearity at the shoulders of the
cylinder. When the water encounters and overlaps at the rear side of cylinder, the run-up
phenomenon can also be observed.

(a) (b)

(c) (d)
Figure 11 Local free surface around the cylinder

Figure 12 shows the dynamic pressure contours on the cylinder surface at the moment that the
wave crest encounters the cylinder surface for wave at T = 7s and H/L =1/30, 1/16, 1/10. The
dynamic pressure increases with the increasing wave height. For wave at T = 7s and H/L =
1/10 condition, the maximum dynamic pressure in the front of the cylinder can been found.
For each case, the maximum dynamic pressure of the cylinder is close to the free surface.

(a)H/L = 1/30 (b) H/L = 1/16 (c) H/L = 1/10

Figure 12 Dynamic pressure the cylinder

Conclusions
In this paper, numerical simulation of the wave run-up on a fixed surface-piercing cylinder in
regular waves are performed by the in-house naoe-FOAM-SJTU solver. Two wave periods (T
= 7s and 9s) and three wave heights (H/L =1/30, 1/16, 1/10) are conducted to investigate the
wave run-up phenomenon. The predicted RAOs of local surface elevation are compared with
the experimental data and good agreement can be acquired even at the strong nonlinear
interaction location, where secondary crests can be observed for steeper waves. This can be
caused by the overlap between the backward wave from rear part and the incident wave.
Concentric and non-concentric wave fields around the cylinder can be captured by the present
CFD solver. The surface elevation around the cylinder and the dynamic pressure increases
with the increasing wave height. This study shows the capability of the present solver to
investigate the wave run-up on a fixed cylinder. Further work should be focused on high-order
harmonic and detailed flow information around the cylinder.

Acknowledgements
This work is supported by the National Natural Science Foundation of China (51379125,
51490675, 11432009, 51579145), Chang Jiang Scholars Program (T2014099), Shanghai
Excellent Academic Leaders Program (17XD1402300), Program for Professor of Special
Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning (2013022),
Innovative Special Project of Numerical Tank of Ministry of Industry and Information
Technology of China (2016-23/09) and Lloyd’s Register Foundation for doctoral student, to
which the authors are most grateful.

References
[1] Galvin, C. J., Hallermeier, R. J. (1972) Wave runup on vertical cylinders, Coastal Engineering 1, 1955-1974.
[2] Chakrabarti, S. K., Tam, W. A. (1975) Wave height distribution around a vertical cylinder, Journal of Waterways,
Harbours and Coastal Engineering 101, 225-230.
[3] Morris-Thomas, M. T., Thiagarajan, K., Krokstad, J. R. (2002) An experimental investigation of wave steepness and
cylinder slenderness effects on wave run-up, In Proceedings of 21st International Conference on Offshore Mechanics
and Arctic Engineering, OMAE2002-28050.
[4] Nielsen, F. G. (2003) Comparative study on airgap under floating platforms and run-up along platform columns, Marine
Structures 16, 97-134.
[5] Havelock, T. H. (1940) The pressure of water waves upon a fixed obstacle, In Proceedings of the Royal Society of
London A: Mathematical, Physical and Engineering Sciences 175, 409-421.
[6] Hunt, J. N., Baddour, R. E. (1981) The diffraction of a nonlinear progressive wave by a vertical cylinder, Quarterly
Journal of Mechanics and Applied Machematics 34, 69-88.
[7] Trulsen, K., Teigen, P. (2002) Wave scattering around a vertical cylinder: fully nonlinear potential flow calculations
compared with low order perturbation results and experiment, In Proceedings of 21st International Conference on
Offshore Mechanics and Arctic Engineering, OMAE2002-28173．
[8] Teigen, P., Niedzwecki, J. M. (2003) Wave diffraction effects and runup around multicolumn structure, In Proceedings
of 13th International Offshore and Polar Engineering Conference, 137-144.
[9] Morris-Thomas, M. T., Thiagarajan, k. (2004) The run-up on a cylinder in progressive surface gravity waves: harmonic
components, Applied Ocean Research 26, 98-113.
[10] Cao, H. J., Wan, D. C. (2014) Development of multidirectional nonlinear numerical wave tank by naoe-FOAM-SJTU
solver. International Journal of Ocean System Engineering 4, 52-59.
[11] Cao, H. J., Wan, D. C. (2017) Benchmark computations of wave run-up on single cylinder and four cylinders by naoe-
FOAM-SJTU solver, Applied Ocean Research 65, 327-337.
[12] Cao, H. J., Wan, D. C. (2015) RANS-VOF solver for solitary wave run-up on a circular cylinder, China Ocean
Engineering 29, 183-196.
[13] Danmeier, D. G., Seah, R. K., Finnigan, T., Roddier, D., Aubault, A., Vache, M., Imamura, J. T. (2008) Validation of
wave run-up calculation methods for a gravity based structure. In Proceedings of the ASME 27th International
Conference on Offshore Mechanics and Arctic Engineering, OMAE2008-57625.
[14] Sun, L., Zang, J., Chen, L., Taylor, R. E., Taylor, P. H. (2016) Regular waves onto a truncated circular column: A
comparison of experiments and simulations, Applied Ocean Research 59, 650-662.
[15] Kristiansen, T., Baarholm, R., & Stansberg, C. T. (2004) Validation of second-order analysis in predicting diffracted
wave elevation around a vertical circular cylinder, In Proceedings of 14th International Offshore and Polar Engineering
Conference, 342-349.
[16] Shen, Z. R., Wan, D. C., Carrica, P. M. (2015) Dynamic overset grids in OpenFOAM with application to KCS self-
propulsion and maneuvering. Ocean Engineering 108, 287-306.
[17] Wang, J. H., Wan, D. C. (2016) Numerical simulation of pure yaw motion using dynamic overset grid technology.
Chinese Journal of Hydrodynamics 31, 567-574.
[18] Ye, H. X., Wan, D. C. (2017) Benchmark computations for flows around a stationary cylinder with high Reynolds
numbers by RANS-overset grid approach, Applied Ocean Research 65, 315-326.
[19] Shen, Z. R., Wan, D. C. (2016) An irregular wave generating approach based on naoe-FOAM-SJTU solver, China
Ocean Engineering 30, 177-192.
[20] Liu, Z. H., Wan, D. C. (2016) Numerical simulation of green water on S-175 containership, In Proceedings of the
Second Conference of Global Chinese Scholars on Hydrodynamics, 815-821.
[21] Hirt, C. W., Nichols, B. D. (1981) Volume of fluid (VOF) method for the dynamics of free boundaries. Journal of
Computational Physics 39, 201-225.
[22] Swan, C., Sheikh, R. (2015) The interaction between steep waves and a surface-piercing column, Philosophical
Transactions of the Royal Society of London A 373, 20140114.