MODELLING WAVE INTERACTIONS WITH A
SURFACE-PIERCING VERTICAL CYLINDER USING OpenFOAM
L.F. Chen1, G.C.J. Morgan1, J. Zang1, A. Hillis1 and P.H. Taylor2
1
Department of Architecture and Civil Engineering, University of Bath
2
Department of Engineering Science, University of Oxford
[email protected] [email protected] [email protected] a,
[email protected] [email protected]INTRODUCTION
With the intensification of the global energy crisis, the development of marine renewable
energy is attracting increasing attention. To achieve the renewable energy targets, a rigorous
approach is required leading to the better design of wave energy converters with increased
efficiency. Wave loading is a key factor to be considered for structural design. Either over-
predicted or under-predicted loading would cause severe consequences. Over-predicting the
loading will lead to overdesign and then very expensive structures. The consequence of
under-prediction may even more dangerous and expensive - leading to under-design and
structural failure.
The use of Computational Fluid Dynamics (CFD) codes is becoming increasing important in
engineering design work. The free, open-source library for continuum-mechanics problems
known as OpenFOAM is suitable for solving complex free-surface motions (Weller et al.,
1998). It has been applied to coastal engineering successfully by Morgan et al. (2010, 2011).
The experimental results for the propagation of monochromatic waves over a submerged bar
have been reproduced in his numerical simulations, with up to 6th order harmonics correctly
modelled.
This research is focussed on the assessment of how OpenFOAM performs when applied to
non-linear wave interaction with a vertical surface piercing cylinder, a typical representation
of an offshore wind turbine foundation. A few new functions have been developed to advance
wave generation and wave absorbing capabilities of the model. A series of experiments
performed in the Danish Hydraulic Institute’s shallow water basin in 2009 have been
reproduced using OpenFOAM to test the accuracy of the model predicting wave-structure
interaction problems. The decomposition of the measured signals into harmonics has also
been carried out to examine the effects of wave and loading nonlinearity.
VALIDATION AND RESULT DISCUSSIONS
The experiments were performed at the Danish Hydraulic Institute, details were given in
Zang et al (2010). The shallow water basin (35m x 25m) was used for the tests with a water
depth of 0.505m. A cylinder of diameter 0.25m was suspended from a rigid frame, leaving
only a 1mm gap beneath to the bed of the basin. The total horizontal hydrodynamic force on
the cylinder was measured via a load cell, and 19 wave gauges were placed to monitor the
wave-field around the cylinder. The layout for the wave gauges can be seen in Figure 1. The
numerical results for the free surface elevations at wave gauges of 1, 5, 9, 13, 17 and 19, and
the horizontal force on the cylinder have been compared with experimental measurements for
several cases, of which two cases will be discussed in this paper. These two cases correspond
to the same slenderness (ka=0.37) and same kD (1.39) for the cylinder, but have different
steepness (kA1=0.1, and kA2=0.2). Here k is wave number, a is cylinder radius, D is water
depth and A is wave crest with A1=0.035m, A2=0.07m. For both cases, wave period is 1.22s.
Figure 1: The arrangement of wave gauges in the wave tank
In the numerical simulations, a few key factors have been examined to ensure accurate
reproduction of the experiments. As the wave generation in the numerical model is via the
flux into the computational domain through a fixed vertical wall, different from wave
generation by piston paddles in the experiments, a careful match between experiments and
numerical models for the first wave gauge (0.77m from the paddle) was made to ensure the
incoming waves in the numerical tank are very close to the waves generated in the
experiments. Wave focus time, focus distance and phase angle are all carefully chosen to
match the experiments. A damping zone is also added in the computational domain to avoid
reflection at the downstream end of the numerical tank.
Time series of experimental and numerical free surface elevations at a few locations along the
central line for the cases with cylinder in place are given in Figure 2. Note that the wave
gauge 19 is placed close to the side wall of the numerical tank for a second check of the
incoming waves. At location of wave gauge 9, which is 2mm in front of the upstream
stagnation point of the cylinder, all wave components are in phase and the wave group
produces a large energetic events. Wave breaking at wave gauge 13, which was observed in
experiments, may lead to a relatively larger discrepancy between the predicted and measured
values of the free surface elevations. A finer mesh would be required to obtain a better match.
Apart from gauge 13, the numerical model appears to have captured all the main physical
features of the nonlinear focused wave interaction with the vertical cylinder, with close
matching of both crest values of the free surface and horizontal forces, and the wave shapes.
In order to extract the harmonic structure of the free surface and wave loading on the cylinder,
both crest focused waves and trough focused waves were performed, both in the experiments
and the numerical simulations. Following Zang et al. (2006), Zang et al. (2010) and
Fitzgerald et al. (2012), our simple phase-based separation method has been applied to
decompose higher order harmonics cleanly without cross-contamination between adjacent
harmonics. Table 1 shows the comparisons of the crest values of each harmonic of free
surface elevations and horizontal forces for wave amplitudes with the cylinder in place. This
is clear that the 2nd order components for case 2 are much larger than case 1, about 4-5 times
larger than the values in case 1, which is consistent with the expected Stokes-type behavior.
Figure 2: Time series of free surface elevations and horizontal forces for both cases with
cylinder in place. The experimental data is in thick grey line and the numerical results are in
thin black line.
Given the good match for all the components, we are confident that accurate predictions of
the free surface elevations in the wave-field and the horizontal force on the cylinder can be
made by using the present model based on OpenFOAM.
Peak free surface elevations (m) Peak horizontal forces (N)
case 1 case 2 case 1 case 2
harmonics
kA1=0.1 kA2=0.2 kA1=0.1 kA2=0.2
Exp Num Exp Num Exp Num Exp Num
long wave 0.0016 0.0017 0.0050 0.0130 1.91 1.80 7.50 7.50
linear 0.0530 0.0560 0.0930 0.1050 40.0 32.0 67.0 57.0
2nd 0.0105 0.0104 0.0370 0.0430 2.20 3.20 9.50 12
3rd 0.0020 0.0020 0.0080 0.0130 1.30 0.70 9.50 4.50
4th 0.0055 0.0075 0.0057 0.0034 1.00 0.20 1.70 2.70
Table 1: Each harmonic of free surface elevations at the location of WG9 and horizontal
forces for both cases with cylinder in place. Exp means Experimental data and Num means
Numerical results.
From the preliminary results, we conclude that the present model based on OpenFOAM can
accurately predict the non-linear wave interaction with a vertical cylinder up to at least the 4th
frequency harmonic. By using the crest-trough phase-based separation method, we can
reproduce harmonic structure in both the free surface and the wave loading on the structure.
Further examples of the validation and the harmonic structure of the free surface and
hydrodynamic loading on a surface-piercing column will be presented at the workshop.
ACKNOWLEDGEMENTS
The first author acknowledges the financial support of the University of Bath for her PhD
study. The second author thanks the Great Western Research, Edenvale Young Associates
Ltd and the University of Bath for the support of his PhD study.
REFERENCES
1. Weller, H. G.; Tabor, G.; Jasak, H. & Fureby, C. (1998) A tensorial approach to computational
continuum mechanics using objectoriented techniques. Computers in Physics, 12(6): 620–631.
2. Morgan, G. and Zang, J. et al. (2010) Using the rasInterFoam CFD model for wave transformation
and coastal modelling. The 32nd international Conference on Coastal Engineering (ICCE 2010).
3. Morgan, G. and Zang, J. (2011). Application of OpenFOAM to Coastal and Offshore Modelling. In:
The 26th International Workshop on Water Waves and Floating Bodies, Greece.
4. Futzgerakd, C. Grice, J. and Taylor, P.H. et al.(2012) Phase manipulation and the harmonic
components of ringing forces on a surface-piercing column. Proc. 27th IWWWFB, Denmark.
5. Zang, J., Taylor, P.H. and Tello, M. et al. (2010). Steep Wave and Breaking Wave Impact On Offshore
Wind Turbine Foundations -- Ringing Revisited. Proc. 25th IWWWFB, China.
6. Zang, J., Gibson, R., Taylor, P. H., Eatock Taylor, R. and Swan, C. (2006). Second order wave
diffraction around a fixed ship-shaped body in unidirectional steep waves. Journal of Offshore
Mechanics and Arctic Engineering, 128 (2), pp. 89-99.
