POLISH MARITIME RESEARCH 1 (109) 2021 Vol. 28; pp.

42-47
10.2478/pomr-2021-0004

HEAVE MOTION OF A VERTICAL CYLINDER

WITH HEAVE PLATES

Ewelina Ciba
Gdańsk University of Technology, Poland

ABSTRACT

The shape of a vertical cylinder resembles the classic form of a spar platform. Spar platforms are floating platforms
that are successfully used in waters of great depths and have several advantages that mean they are readily used in
the oil industry. Many of these advantages are also relevant to their application for offshore wind turbines, which is
currently being considered. In the hydrodynamic analysis of spar platforms, the determination of their hydrodynamic
coefficients plays an important role. They can be determined based on the free decay test. The study presents a method
for determining the hydrodynamic coefficients of an object based on the free decay test. The results of free oscillation
calculations with the help of numerical fluid mechanics tools are presented and compared with the results of the
experiment and analytical solution. The application of determined coefficients and their significance for floating
platforms are discussed. The influence of change in the form of an additional damping element on the behaviour of
spar structures is shown.

Keywords: spar platforms, heave plates, damping coefficient, added mass coefficient

INTRODUCTION This solution has become very popular and consistent results
of analyses of its effectiveness can be found in the literature.
A spar is a type of floating platform designed for great The increase in the added mass, the damping coefficient and
depths. The structure of a spar platform is a vertical cylinder the related increase of the natural period under the influence
with positive buoyancy with an appropriate system of of the heave plate have also been confirmed [5, 6, 7].
anchoring the platform to the seabed. They are always stable Conventional offshore structure behavior calculations in
because their centre of buoyancy is above the centre of gravity. offshore conditions are based on the potential theory and
They can be used as mobile drilling rigs and their cable system do not take into account viscous damping. This is a good
(catenary) is easy to install, use and carry [1]. approximation, but in some cases the viscosity should be
However, with long wave periods, the phenomenon of included. Especially for the near resonance frequency [8]. The
resonance may appear, which is largely influenced by damping importance of viscosity also increases with the presence of
related to water viscosity [2, 3]. In [4], several alternative hull edges in the structure, which was demonstrated by comparing
shapes that improve the damping in the vertical movement of the behavior of the hemispherical, rounded base and flat base
the structure are presented. The best of the presented solutions is structures in [9]
a heave plate, a damping plate attached to the floating structure The resulting vortices and their impact on the drag
that has a distinct advantage in reducing the motion response of coefficient depending on the geometry of the heave plate
a floating structure by increasing the added mass and damping. are shown e.g. in [10] or [11].

42 POLISH MARITIME RESEARCH, No 1/2021

 A good technique is to supplement linear theories with [16, 5]. As a result, considerations regarding spar platforms
viscous drag coefficients obtained from experimental studies are becoming very popular again. Due to the different nature
or calculations using the RANSE (Reynolds-Averaged Navier of the forces acting in the case of offshore wind platforms than
Stokes) model with turbulence models. This approach is in the case of oil rigs, there are also new problems with the
described in [9] and [8]. The possibility to consider the effect dynamics of these platforms related to the action of wind force.
of viscosity is also available in the commercial Ansys Aqwa This study presents the calculation results of the numerical
software. Software uses three dimensional panel method to fluid dynamics of the free decay test for a smooth cylinder
solve linearized hydrodynamic radioation and diffraction and a cylinder with a heave plate (Fig. 1), and compares them
problem in frequency domain [12] with the results of analytical calculations and experiment.
Potential flow solutions with added viscosity factors allow The damping coefficients and the added mass coefficient
you to get good results in a reasonable amount of time. Time- are determined and compared with the values presented in
consuming RANSE calculations are performed only once to the literature. The confirmation of the correctness of the
calculate the coefficients, and then a series of analyzes can be applied method allows us to state that the computational fluid
performed for different conditions and configurations with dynamics (CFD) calculations can be used to correctly estimate
faster models based on the potential theory, extended by the the hydrodynamic coefficients of the analysed platforms. This
coefficients mentioned. is of great importance in the process of designing a structure,
In [13], modelling of the spar platform dynamics with the as it allows to carry out extensive parametric analyses at a low
use of drag coefficient, inertia coefficient and added mass cost and hence to choose the best solution, which should then
coefficient was presented. In addition, in [10], the added mass be checked using model tests. Moreover, CFD methods allow
and damping coefficients related to the plate viscous damping for analyses of various shapes and configurations, which is
were presented. One of the methods of estimating the viscous not always possible with analytical methods.
damping is to conduct free decay tests, as discussed in [7].
A very interesting and extensive analysis of the heaving
plate impact using the CFD method is presented in [11] MATHEMATICAL DESCRIPTION
Detailed parametric studies were conducted numerically on OF THE PROBLEM
spar with single and two heave plates. Effects of scaling ratio,
diameter ratio, heave plate position to diameter of heave plate Submerged, under the influence of the initial force, the
ratio, heave plate position to draft ratio, relative spacing and cylinder will then make an oscillating motion relative to the
spacing between heave plates to draft ratio were examinated. equilibrium position, with the amplitude of the movement
As a result, the ranges of the analyzed parameters were given decreasing with time. A mathematical description of the
for which the heave motion reduction is the greatest. problem can be found in the literature [17].
In recent years, there has been a dynamic development of We assume that the moving body is a rigid body and we
offshore wind energy. There are concepts of wind turbines replace surface and volume forces by the movement of the
based on various types of supporting structures. The concept point associated with the origin of the coordinate system. We
of a turbine based on a jack-up platform is described in [14]. also assume that the coefficients of the equation are constant
A concept based on a tension leg platform was presented over time, which causes the equation to become a linear. This
in [15]. Most often, however, due to the large depths of the assumption is only true for a small amplitude of motion. In
planned installation areas, the concepts of offshore wind this case, the deflection is not more than 6 cm so they can be
turbines are based on spar-type platforms, as presented in used. The cylinder motion equation is given as:

Fig. 1. Geometry of tested cylinders: a) smooth; b) with a heave plate

POLISH MARITIME RESEARCH, No 1/2021 43

 (m + a) · + b · ż + c · z = 0 (1)

where:
z – vertical displacement [m]
m – solid mass of cylinder [kg]
a – hydrodynamic mass coefficient [kg]
b – hydrodynamic damping coefficient [kg/s]
c – restoring spring coefficient [kg/s2]

Equation (1) can be written as:

+ 2v · ż + ω02 · z = 0 (2)

where the damping coefficient and the undamped natural

frequency are defined as:

(3) Fig. 2. Schematic of computational domain

A non-dimensional damping coefficient κ is defined as:

(4)

By knowing the results of the free decay tests, we can calculate:

(5)

Depending on the averaged displacement amplitude:

Fig. 3. Heave plate geometry
(6)
The assumed initial conditions were as follows: the share
We calculate the hydrodynamic added mass coefficient as: of water and air volume, and the initial pressure distribution
of 1 atm above the free surface and hydrostatic pressure
(7) below the free surface. The cylinder was treated as a rigid
body with one degree of freedom. The mass is equal to the
In the case of initial displacement za, Eq. (2) takes the form: displacement at draught 0.3 m, i.e. 6 kg. The initial solid
velocity was assumed to force the initial cylinder displacment.
(8) In the case of a smooth cylinder, the mesh was compacted
around the cylinder (0.005 m) with a 0.002 m thick wall layer
Due to the fact that for the frequency of free oscillations and two layers, by its expected displacement, and in the area
ωz2 = ω 02 · v2 and damping is small v < 0.20, v2 << ω2, we can of the free surface (0.01 m). The grid in the plane of symmetry
skip v2 and we can write that ωz ≈ ω 0. with the division into moving and fixed parts is shown in
Fig. 4. The number of grid elements was over 6 million.

RANSE CFD CALCULATIONS

CFD calculations were made using STAR-CCM+. Unsteady

calculations were performed in the three-dimensional
domain, using the volume of fluid and K-epsilon turbulence
models. Cylinder displacements were modelled using the
dynamic fluid body interaction module using an overset mesh.
Each of the boundary walls was given a no-slip wall condition.
The computational domain was prepared as in Fig. 2.
Fig. 4. Calculation grid in the plane of symmetry for smooth cylinder
The shape of the second cylinder was changed by a heave
plate with diameter DHP = 0.178 m and thickness t = 0.005 m. In the case of a cylinder with an additional heave plate, the
The transition between the edge of the ring and the cylinder mesh was refinement in the plate area (0.002 m), the thickness
wall is rounded at radius r = 0.02 m. The geometry of the of the prism layer was reduced to 0.0015 m and the number
cylinder with a heave plate is shown in Fig. 3. of layers was increased to seven. The mesh for the cylinder

44 POLISH MARITIME RESEARCH, No 1/2021

with the plate is shown in Fig. 5. The number of elements was
increased to over 7 million.

Fig. 7. Vertical cylinder acceleration measured for an example series of tests

Fig. 5. Calculation grid for a cylinder with a heave plate
Based on the measured accelerations, a displacement
Calculations were performed for both cases with time plot was created knowing that the amplitude is equal to
step t k = 0.01 s. the quotient of the acceleration and the square of the free
vibration frequency at the points of maximum deflection.

COMPARISON OF CALCULATION (9)

RESULTS WITH EXPERIMENT
The comparison of the displacement diagrams for both
In order to validate the calculations, an experiment was cases is shown in Fig. 8 for a smooth cylinder and in Fig. 9
conducted whose conditions were equivalent. The same for a cylinder with a heave plate.
mass and geometry were used, and the resulting draft of the
structure. The oscillation was forced by the application of an
external force, the effect of which then ceased. The force was
selected in such a way that the first free displacement was the
same as in the performed calculations. The experiment was
carried out in such a way that the weight of the appropriate
mass and the accelerometer were placed inside the empty
cylinder of the set geometry. Then, an initial draft greater than
the one resulting from the displacement balance was forced
and released, recording the accelerations of the oscillating
cylinder. It was assumed that the cylinder only moves
vertically. Its slight deviations have been neglected. Two series
of tests were performed for each case. A photograph from the
study is shown in Fig. 6.
Fig. 8. Heaving a smooth cylinder

Fig. 9. Heaving a cylinder with a heave plate

Fig. 6. Model tests of a damping ring cylinder
To compare both cases, the non-dimensional damping
As a result of the measurements, the acceleration changes coefficient κ (Eq. (5)) is shown in the plot depending on the
over time were obtained. The measured values for one of the average amplitude (Eq. (6)).
test series are shown in Fig. 7.

POLISH MARITIME RESEARCH, No 1/2021 45

 differences in the dimensionless attenuation coefficient are
not greater than the differences between the values obtained
from both series of the experiment (fig. 11 and 12). The results
of the calculations and the experiment differ for greater
displacement amplitudes. Perhaps the calculations should
be repeated with a shorter time step to improve convergence
for these points.
As shown in the diagram (Fig. 10), the value of the damping
coefficient depends on the displacement amplitude, and
therefore the cylinder velocity. The conclusion is that when
extending the calculations based on the potential method with
coefficients related to viscosity, either the average damping
coefficient for the anticipated speed range should be used
Fig. 10. A non-dimensional damping coefficient
or, if possible, the value of the coefficient as a function of
plot depending on the average displacement amplitude velocity should be used.
Comparing the results for the cases of a smooth cylinder and
The natural frequency ω 0 (Fig. 11) and the added mass a cylinder with an additional heave plate, even a slight change
coefficient a (Fig. 12) are also compared. in shape whilst maintaining the same weight of the structure
causes an increase in damping. An additional element also
reduces the frequency of oscillations and causes an increase
in the mass of the accompanying water. The obtained values
of the coefficients are similar to those obtained by other
methods presented in the literature. A further conclusion
can be drawn from this that when designing the geometry of
a spar platform, it is worth considering the use of additional
damping elements of this type.
Confirmation of the correct application of CFD methods
for the analysis of hydrodynamic coefficients is of great
importance for the analysis of various concepts of additional
Fig. 11. Natural frequency damping elements for structures for offshore wind turbines.
This will allow, at relatively low costs, for the analysis of many
different solutions and we can select the best of them for
model tests necessary to verify the obtained results. It is
can be used to carry out calculations for plates with jagged
edges, equipped with movable elements, and for flexible
plates. CDF analyses allow us to carry out many series of
calculations depending on changes in individual parameters
of the elements (e.g. increasing the diameter of the plate) that
will also help to trace the dependence of the coefficients on
the selected parameters.
Fig. 12. Added mass coefficient
ACKNOWLEDGEMENT
SUMMARY
The calculations were performed using a Simcenter STAR-
The concepts of emerging wind platforms are often based CCM+ license.
on spar platforms and their modifications. This increases the
interest in these platforms and the need for numerical analyses
for the designed structures. Due to the time consuming REFERENCES
nature of full RANSE CFD calculations, calculations by
hybrid methods are a  good solution, using much faster 1. Jain A.K., Agarwal A.K. (2003): Dynamic Analysis of
calculation methods based on hydrodynamic coefficients Offshore Spar Platforms. Defence Science Journal, Vol. 53,
determined by RANSE CFD analyses and model tests. This No. 2, pp. 211–219
study presents a method for determining such coefficients.
Based on the results presented, it can be concluded that CFD 2. Sharman K. T., Robertson A., Lewis J. (2019): Heave Plate
calculations allow us to reliably map reality. In both cases, Hydrodynamics for Offshore Wind Turbine Applications.
the large convergence of CFD calculation results, analytical EERA DeepWind, Trondheim
solutions and experiment results was obtained, and the

46 POLISH MARITIME RESEARCH, No 1/2021

3. Liu Y. Yan H. Yung T. W. (2010): Nonlinear Resonant of Cell Spar Type During its Installation. Polish Maritime
Response of Deep Draft Platforms in Surface Waves. ASME Research, Vol. 26, 4(104), 109–116.
2010 29th International Conference on Ocean, Offshore
and Arctic Engineering 17. Journee J. M. J., Massie W. W. (2001): Offshore
Hydromechanics. Delft University of Technology.
4. Bhatta D. D. (2007): Computation of Added Mass and
Damping Coefficients due to a Heaving Cylinder. J. Appl. 18. Kim et al. M.-H. (2012): Spar platforms: Technology and
Math. Comput., 23, 127 140 Analysis Methods. American Society of Civil Engineers.

5. Gilloteaux J. Ch., Bozonnet P. (2014): Parametric Analysis of

a Cylinder Like Shape Floating Platform Dedicated to Multi-
Mega Watt Wind Turbine. IFP Energies Nouvelles, Applied
Mechanics division Rueil-Malmaison, France.
CONTACT WITH THE AUTHOR
6. Fisher F. J., Gopalkrishnan R. (1998): Some Observations on
the Heave Behavior of Spar Platforms. Journal of Offshore
Mechanics and Arctic Engineering. Ewelina Ciba
e-mail: [email protected]
7. Koh H. J., Cho I. J. (2016): Heave Motion Response of
a Circular Cylinder with the Dual Damping Plates. Ocean Gdańsk University of Technology
Engineering, 125, 95–102. Narutowicza 11/12
80-233 Gdańsk
8. Zhao Ch., Cao F., Shi H. (2020): Optimisation of heaving Poland
buoy wave energy converter using a combined numerical
model Applied Ocean Research 102 (2020) 102208

9. Gu, H., Stansby, P., Stallard, T., & Carpintero Moreno, E.

(2018). Drag, added mass and radiation damping ofoscillating
vertical cylindrical bodies in heave and surge in still water.
JOURNAL OF FLUIDS AND STRUCTURES, 82, 343–356.
https://doi.org/10.1016/j.jfluidstructs.2018.06.012

10. Tao, L., Cai, S. (2004): Heave Motion Suppression of a Spar

with a Heave Plate. Ocean Engineering.

11. A. Subbulakshmi, R. Sundaravadivelu (2016) Heave damping

of spar platform for offshore wind turbine with heave plate
Ocean Engineering 121 (2016) 24–36

12. S. Sudhakar, S. Nallayarasu (2011) Influence of Heave Plate

on Hydrodynamic Response of Spar OMAE2011-49565

13. Haslum, H. A. (2000): Simplified Methods Applied to

Nonlinear Motion of Spar Platforms. Norwegian University
of Science and Technology, Trondheim.

14. Dymarski P. (2019): Design of Jack-up Platform for 6 MW

Wind Turbine: Parametric Analysis Based Dimensioning
of Platform Legs. Polish Maritime Research, 26, 183–197.

15. Żywicki J., Dymarski P., Ciba E., Dymarski C. (2017): Design
of Structure of Tension Leg Platform for 6 MW Offshore Wind
Turbine Based On Fem Analysis. Polish Maritime Research,
Vol.–24, s1, 230–241

16. Dymarski P., Dymarski C., Ciba E. (2019): Stability Analysis

of the Floating Offshore Wind Turbine Support Structure

POLISH MARITIME RESEARCH, No 1/2021 47