Chapter 20

Modeling of Surf-Riding in Irregular

Waves

Vadim L. Belenky, Kostas J. Spyrou and Kenneth M. Weems

Abstract Surf-riding is an important phenomenon for the evaluation of ship

dynamic stability, as it is related to one of the principal mechanisms of broaching-to.
The evaluation of the probability of surf-riding in irregular waves is a necessary step
toward determining the probability of broaching-to following surf-riding. To facili-
tate the probabilistic study of surf-riding, a simple model of surging and surf-riding
in irregular waves of variable bandwidth is introduced. This model can be used to
identify patterns of surf-riding in irregular waves.

Keywords Surf-riding · Irregular waves

20.1 Introduction

The dynamics of broaching-to originating from surf-riding in regular waves is now

well understood from a global dynamical systems’ viewpoint (Spyrou 1996). A
cornerstone of this phenomenon is the appearance of a pair of surf-riding equilibria
of which the one that attracts in surge may be a repeller in yaw, depending on the
effectiveness of rudder control. Another key issue is the possible dominance of this
equilibrium in state space due to a “homoclinic connection” bifurcation that renders
surf-riding inevitable.
The physical mechanism of surf-riding includes the appearance of dynamical
equilibria and a ship’s attraction to the stable equilibrium (Kan 1990). The equilibria
appear when the wave surging force becomes large enough to offset the difference
between the ship’s thrust and resistance at speed equal to wave celerity. The equi-

V. L. Belenky (B) · K. M. Weems

David Taylor Model Basin (NSWCCD), West Bethesda, MD, USA
e-mail: vadim.belenky@navy.mil
K. J. Spyrou
School of Naval Architecture and Marine Engineering, National Technical University of Athens,
Athens, Greece

© Springer Nature Switzerland AG 2019 347

V. L. Belenky et al. (eds.), Contemporary Ideas on Ship Stability, Fluid Mechanics
and Its Applications 119, https://doi.org/10.1007/978-3-030-00516-0_20
348 V. L. Belenky et al.

Fig. 20.1 On the appearance Longitudinal forces

of dynamic equilibria
Positions on the wave

Wave phase

Balance between
thrust and resistance

Unstable
equilibrium Stable
equilibrium

librium points are the positions of the ship on the wave where the forces balance
exactly.
To illustrate this, consider surf-riding in regular waves and plot the wave-induced
surging force as a function of the ship’s position on the waves, taken with the opposite
sign; see Fig. 20.1. In this plot, the horizontal axis is the position of the ship’s center of
gravity ahead of the wave crest, the dashed line is the wave profile, and the solid line
is the wave surging force, with a negative value indicating a forward (accelerating)
force.
Since the commanded speed is insufficient to propel the ship at wave celerity in
calm water, additional wave force is necessary to drive the ship at wave celerity. If the
amplitude of the wave surging force exceeds the absolute value of the balance between
thrust and resistance, two intersection points appear, as shown in Fig. 20.1. Those
will be called “surf-riding equilibria” (knowing that this is not an exact condition
of equilibrium); one shows stability features (black point, located around the wave
trough) and the other behaves as unstable (empty point, located around wave crest).
While these considerations are well-established in the field, they are repeated here
in order to highlight the difference between regular and irregular waves. If Fig. 20.1
is considered as a snapshot of forces in an irregular wave, all of the elements of
the surf-riding problem can be transferred from regular waves to irregular waves.
However, as these elements may vary in time and space, the problem should be
defined simultaneously in space and time—creating a spatial-temporal framework
for surf-riding in irregular seas, see Fig. 20.2.
As the wave surging force and celerity change with the randomness of the irregular
wave, these dynamic equilibria appear and disappear and move with unsteady speed.
As a result, these equilibria points cannot be solutions of the equations of motion
and, strictly speaking, do not have all the properties of equilibria in a physical sense.
Nevertheless, they still define the topology of phase space. This brief argument
indicates a significant difference between surf-riding and broaching in regular and
irregular waves.
20 Modeling of Surf-Riding in Irregular Waves 349

Fig. 20.2 Random Time, s

surf-riding in 50
spatial-temporal framework

40

30

20

Unstable
10 equilibria Stable
equilibria

0 200 400 600

Longitudinal position, ξG, m

While the overview of this problem is available from Belenky et al. (2016), the
present text is focused on the simplest mathematical model for surf-riding in irregular
waves, originally described in Belenky et al. (2011, 2012).

20.2 Surging in Irregular Waves

Consider a simple model for one-degree-of-freedom nonlinear surging:

   
(M + A11 )ξ̈G + R ξ̇G − T ξ̇G , n + FX (t, ξG )  0 (20.1)

Here M is the mass of the ship, A11 is the longitudinal added mass, R is the
resistance in calm water, T is the thrust in calm water, n is the number of propeller
revolutions, F X is the surging component of the Froude-Krylov wave force, and ξG is
the longitudinal position of the center of gravity in the Earth-fixed coordinate system.
The dot above the symbol stands for temporal derivative. The following polynomial
presentations are used for the resistance and thrust in calm water (for compatibility
with Spyrou 2006):
350 V. L. Belenky et al.

R(U )  r1 U + r2 U 2 + r3 U 3
T (U, n)  τ1 n 2 + τ2 nU + τ3 U 2 (20.2)

The irregular waves are presented as a spatial–temporal stochastic process using

the standard Longuet-Higgins model, based on the linear dispersion relation:


N
ζW (t, ξ )  ai cos(ki ξ − ωi t + ϕi ) (20.3)
i1

where ai is the amplitude, k i is the wave number, ωi is the frequency, ξ is a coordinate

in the direction of wave propagation, and ϕi is a random, uniformly distributed phase
shift of the wave component i. As the model is meant at this stage to be qualitative,
a linear wave-body formulation is appropriate. Therefore


N
FX (t, ξG )  A Xi cos(ki ξ − ωi t + ϕi + γi ) (20.4)
i1

Since a body-linear formulation is adopted, the amplitude AXi and phase shift γi
are available from response amplitude and phase operators:

A Xi  ai R AO(ki ) (20.5)

⎛ ⎞2 ⎛ 0.5L ⎞2

0.5L


R AO(ki )  ρgki 
⎝ C(x, ki ) cos(ki x)d x ⎠ + ⎝ C(x, ki ) sin(ki x)d x ⎠
−0.5L −0.5L
(20.6)

0
C(x, ki )  2 exp(ki z)b(x, z + d)dz (20.7)
−d

where x and z are measured in the ship fixed coordinate system (positive forwards of
amidships and upward from the base line), b(x, z) is the molded local half-breadth,
ρ is a mass density of water, L is a length of the ship, and d is the amidships section
draft. Figure 20.3 shows the RAO of the surging wave force for the tumblehome ship
from the ONR topside series (Bishop et al. 2005). The phase shift γi is presented as
Fig. 20.4.
 0.5L 
−0.5L C(x, ki ) sin(ki x)d x
γi  arctan  0.5L (20.8)
−0.5L C(x, ki ) cos(ki x)d x
20 Modeling of Surf-Riding in Irregular Waves 351

Fig. 20.3 RAO of surging RAO(k), kN/m

force
1500

1000

500

k, m-1
0 0.05 0.1 0.15 0.2 0.25

Fig. 20.4 Phase shift of γ, rad

surging force 4

k, m-1

0 0.05 0.1 0.15 0.2 0.25

-2

-4

20.3 Spectral Bandwidth: Modeling and Influence

To observe the appearance and development of the qualitative difference between

surf-riding in regular and irregular seas, it may be useful to start from a spectrum
with extremely narrow bandwidth. This idea has been implemented in the form of a
“bilinear filter.”
⎧
⎪
⎪ 0 ω < blow ω
⎪
⎪
⎪
⎨ ω−ωm −blow ω
blow ω ≤ ω < ωm
blow ω
F(ωi )  (20.9)
⎪
⎪
ω−ωm +bup ω
− blow ωm ≤ ω ≤ bup ω
⎪
⎪ ω
⎪
⎩
0 ω > bup ω

where ωm is the modal frequency of the spectrum while ω is the frequency step.
The filter consists of two lines: the low frequency corresponds to the index blow and
the high frequency index is bup . These two indices are parameters for controlling
the spectrum bandwidth. To keep the variance of the wave elevation constant, a
normalization coefficient is used:
352 V. L. Belenky et al.

Fig. 20.5 Example of ai , m

changing wave spectrum 0.6
“Filtered”
bandwidth with bilinear filter

0.4

Original
0.2

ωi, 1/s
0.4 0.6 0.8 1 1.2

N 2
i1 ai
K N  N 2
(20.10)
i1 ai F(ωi )

A sample result of the bilinear filter is shown in Fig. 20.5. After discretizing a
Bretschneider spectrum with 174 frequencies, a total filtered spectrum is created
by selecting the lower boundary 10 frequencies below the modal frequency and the
upper boundary 20 frequencies above the modal frequency. This corresponds to a
decrease of the spectrum bandwidth parameter from 0.703 to 0.21.
The filtering allows regular wave simulations to be carried out in an irregular wave
framework; it is enough to set the indices of the high and low frequency boundaries
to the modal frequency and only one component will remain in Formula (20.3). The
random phase in Formula (20.3) allows the effect of initial conditions to be observed.
It can be clearly seen in Fig. 20.6, which plots the time history of the surge velocity for
wave spectrum and speed setting but with different wave phases. The speed setting
refers to the ship’s calm water speed for the specified constant propeller rate, which is
also the initial speed in the simulation. For this regular wave case with a speed setting
of 27.5 kn (14.1 m/s), there is a co-existence of surging (Fig. 20.6a) and surf-riding
(Fig. 20.6b) which can be realized through different initial phases.
Adding just one additional frequency significantly changes the picture. As pointed
out in Spyrou et al. (2016, 2018), the dynamic behavior becomes very rich—it
includes oscillatory surf-riding, period doubling and chaotic motions. The type of
observed dynamical behavior is dependent on how far the frequencies are apart and
the ratio of the amplitudes.
Figure 20.7 shows the wave elevation with three frequencies. As the frequencies
are very close to each other (see insert in Fig. 20.7), the waves are almost harmonic
with slowly changing amplitude. Figure 20.8 shows the ship response. It starts with
acceleration to 19 m/s and apparent surf-riding until about 300 s when it turns to peri-
odic surging. Then there is another acceleration around 600 s, followed by apparent
surf-riding. Thus, the considered tri-chromatic case seems to model the “catch-and-
release” type of surf-riding, expected for the irregular waves, see Fig. 20.2.
20 Modeling of Surf-Riding in Irregular Waves 353

(a)

(b)

Fig. 20.6 Simulation in regular wave—co-existence of surging (a) and surf-riding (b)

1.5 ai, m
1
ζw, m
4 0.5
ωi, 1/s

2 0.4 0.6

-2

-4 0 50 100 150 200 250 t, s

Fig. 20.7 Wave elevation at the origin composed of 3 frequencies

Fig. 20.8 Simulation with 3 frequencies

Figure 20.9 shows the response for the original Bretschneider spectrum. Both
records contain large amplitude asymmetric oscillation, sometimes with flattened
peaks. This asymmetric behavior may be a manifestation of surf-riding in irregular
seas. To prove this, however, one needs to find the positions and evolution of the
surf-riding equilibria.
354 V. L. Belenky et al.

Fig. 20.9 Simulation with (a)

original Bretschneider
spectrum

(b)

20.4 Surf-Riding in Tri-Chromatic Seas

The very definition of celerity in irregular waves is actually a very deep problem
and is given full consideration in the next chapter of the book, as well as in Spyrou
et al. (2012, 2014), which discusses the formulation of a practical definition of wave
celerity and the implementation of schemes for evaluating it for theoretical and
numerical analysis. An extremely simplified version of the approach is used here,
in which the local celerity is defined by identifying the three profile zero-crossing
points that are closest to the ship and tracking their velocities from time step to time
step.
Once this approximate concept of wave celerity is accepted, the next objective
is to see the relation between the visible surf-riding behavior and appearance of
equilibrium. Tri-chromatic case, shown in Figs. 20.7 and 20.8, is chosen for further
study. While still very simplistic, simulations using this wave and the same simplified
ship surging model show three transitions—two captures and one release—over the
passing of six waves. The time histories of the celerities of these six waves are shown
in Fig. 20.10.
Along with a larger (about 1 m/s magnitude) changes in wave celerity, there are
quite dramatic peaks with three secondary peaks on the top. These secondary peaks
may be artifacts of the simplified wave tracking scheme and/or results of waves
overtaking one another.
Figure 20.11a shows the time history of the instantaneous ship speed and the wave
celerity of the “current” wave, the wave closest to the ship, at any given time. One
can see that the celerity curve at Fig. 20.11a is a combination of all six time histories
in Fig. 20.10, and can be discontinuous as the current wave changes.
Figure 20.11a also shows the temporal boundaries for waves. During periodic
surge motion, they almost always coincide with a down-crossing if the commanded
speed line. Figures 20.11b through 20.11q are the spatial snapshot corresponding to
20 Modeling of Surf-Riding in Irregular Waves 355

c(t),m/s

20 Wave #6

18

Wave #5
20

18

Wave #4
20

18

Wave #3
20

18

Wave #2
20

18

20 Wave #1

18
0 100 200 300 400 500 600 700 800 t,s

Fig. 20.10 Time histories of wave celerities

specific instant of time as noted in the captions. These instances of time are identified
in Fig. 20.11a with lettered arrows referring to the respective spatial snapshot.
On each spatial snapshot plot, the blue line shows the wave profile around the
ship, with the horizontal position of the diamond marking the ship’s position relative
to the wave. The circles on the wave profile mark three zero-crossing points that are
tracked to estimate wave celerity, of which the outermost circles define the spatial
boundary of the “current” wave. The direction of the ship (and wave) motion is to the
right. The vertical position of the diamond indicates the ship speed, while the middle
line (zero line of the wave profile) marks the commanded speed in calm water. The
increase of speed is downwards.
Each spatial snapshot also contains a plot for surging wave force (red curve) and
the balance between the available thrust and resistance at the current wave celerity
in the same scale (lower brown line).
The first spatial snapshot, Fig. 20.11b, corresponds to the initial conditions, with
the instantaneous speed equal to the commanded speed. The ship has just encountered
wave #1 and is located just within its boundary. The surf-riding equilibria exist, since
356 V. L. Belenky et al.

Wave #3 Wave #5
Wave #2 Wave #4
(a) Wave #1 Wave #6
25 Vs, m/s
(c) Ship Speed
(e)
(o) Celerity
20
(f) (h) (j) (l) (r)
(d) (g)
15 Commanded speed
(b) (i) (p)
(k) (m) (n) t, s
10
0 100 200 300 400 500 600 700 800 900

(b) t=0 s, wave #1, the (c) t=40 s, wave #1, transition (d) t=74 s, wave #1, transition (e) t=150 s, wave #1, transition
equilibrium exists, to the equilibrium, positive to the equilibrium, negative to the equilibrium has been
attraction has started peak peak completed

(f) t=256 s, wave #1, (g) t=300 s, wave #1 overtakes (h) t=326 s, wave #2 is (i) t=360s, wave #2 overtakes
equilibrium disappears, the the ship, transition to encountered, the previous the ship, the surging force
surging force decreases surging wave has overtaken the ship increases noticeably

(j) t=380 s, wave #3 is (k) t=398 s, wave #3, the (l) t=434 s, wave #4 is (m) t=490 s, wave #4 has
encountered equilibrium has appeared encountered overtaken the ship

(q) t=800 s, wave #6,

(n) t=496 s, wave #5, negative (o) t=530 s, wave #5, passing (p) t=554 s, wave #5 takes transition to the
peak in periodic surging by the unstable equilibrium over the ship equilibrium is completed

Fig. 20.11 Two captures and one release from surf-riding in a three-component irregular wave;
time history (a) and spatial snapshots (b–q)

the surging force crosses the line corresponding to the balance between the thrust
and resistance.
The stable surf-riding equilibrium attracts the dynamical system and one oscilla-
tion period is seen in Fig. 20.11a until approximately t  100 s. The next two spatial
snapshots, Fig. 20.11c, d, correspond to the positive and negative peaks during this
20 Modeling of Surf-Riding in Irregular Waves 357

transition, respectively. The transition is completed and the dynamical system reaches
the stable equilibrium at around t  150 s, in Fig. 20.11e.
Looking at Fig. 20.11b through 20.11e, one can see that the amplitude of the
surging force is decreasing due to a reduction of the wave amplitude. This tendency
leads to the disappearance of the surf-riding equilibria around t  256 s and to the
release of the ship from surf-riding (Fig. 20.11f).
As the ship slows down (Fig. 20.11g), the wave #1 overtakes her, and wave
#2 is encountered at around t  326 s (Fig. 20.11h). The ship experiences the first
almost periodic surge with the positive peak corresponding to the spatial snapshot in
Fig. 20.11i. As expected, wave #2 overtakes the ship quite quickly and wave #3 is
encountered around t  380 s (Fig. 20.11j).
The modulation of wave amplitude and surging force then reverses and they begin
to increase. This is may be already seen in Fig. 20.11h, but becomes quite apparent
in Fig. 20.11i, j. New surf-riding equilibria appear around t  398 s (Fig. 20.11k).
The existence of the surf-riding equilibria has an immediate influence on the
surge motions, which become asymmetric with wider positive peaks and sharper
negative ones (Spyrou 2006). Symmetry is observed during the passing of waves
#4 and #5, during which the surf-riding equilibria exists continuously (Fig. 20.11l
through 20.11o). Figure 20.11o shows how the dynamical system passed near the
unstable surf-riding equilibrium, but the ship is not yet “caught” and wave #5 takes
over (Fig. 20.11p). The ship is finally “caught” by wave #6 and at around t  800
again reaches the stable surf-riding equilibrium (Fig. 20.11q).

20.5 Summary and Conclusions

A simple model of surging and surf-riding in irregular waves has been formulated. A
bilinear filter is introduced to control the bandwidth of the waves, so the connection to
surging/surf-riding response in regular waves can be made. This filter was used for a
visual analysis of the influence of the spectrum bandwidth on the surging/surf-riding
response. An asymmetric response was observed that may incorporate the surf-riding
behavior. The next logical step is to find the locations of the surf-riding equilibria
in time and space. That step has led to the necessity of defining wave celerity in
irregular waves.
The formulation of a viable definition of wave celerity in irregular waves and the
development of robust methods for evaluating it in numerical theoretic analysis is a
quite substantial task and represents its own area of research. For the present study,
a simple definition was adopted, in which the wave celerity was defined as a speed
of zero-crossing points of a wave and evaluated using a point tracking scheme. This
approximate definition of wave celerity in irregular waves nevertheless allowed the
characterization of the surf-riding equilibria (or the lack thereof) for some numerical
case studies involving irregular waves with three components.
358 V. L. Belenky et al.

Acknowledgments The work described in this paper has been funded by the Office of Naval
Research (ONR) under Dr. Patrick Purtell and Dr. Woei-Min Lin, NSWCCD Independent Apply
Research (IAR) program under Dr. Jack Price and by ONR Global under Dr. Richard Vogelsong
and Dr. Salahuddin Ahmed.

References

Belenky, V., Spyrou, K. J., Weems, K. M., 2011, “Split-Time Method for Surf-Riding and Broaching-
To,” Proceedings of 12th International Ship Stability Workshop, (ISSW 2011), Washington DC,
pp. 163–168.
Belenky, V., Spyrou, K.J., Weems, K.M., 2012 “Evaluation of the Probability of Surf-Riding in
Irregular Waves with the Time-Split Method,” Proceedings of 11th International Conference on
Stability of Ships and Ocean Vehicles (STAB 2012), Athens, Greece, pp. 29–37.
Belenky, V., Weems, K. M. Spyrou, K. J., 2016 “Towards a Split-Time Method for Estimation of
Probability of Surf-Riding in Irregular Seas,” Ocean Engineering Vol. 120 pp. 264–273.
Bishop, R. C., W. Belknap, C. Turner, B. Simon, and J. H. Kim, 2005, “Parametric Investigation
on the Influence of GM, Roll damping, and Above-Water Form on the Roll Response of Model
5613”. Report NSWCCD-50-TR-2005/027, Bethesda, Washington DC.
Kan, M., 1990, “Surging of Large Amplitude and Surf-Riding of Ships in Following Seas,” Selected
papers in Naval Architecture and Ocean Engineering, The Society of Naval Architects of Japan,
28, Ship and Ocean Foundation, Tokyo, 14 p.
Spyrou, K.J., 1996, “Dynamic Instability in Quartering Seas: The Behavior of a Ship During Broach-
ing,” Journal of Ship Research, Vol. 40, No. 1, pp. 46–59.
Spyrou, K.J., 2006, “Asymmetric Surging of Ships in Following Seas and its Repercussions for
Safety,” Nonlinear Dynamics, Vol. 43, pp. 149–172.
Spyrou, K.J., Belenky, V., Themelis, N., Weems, K.M., 2012, “Conditions of Surf-riding in an
Irregular Seaway,” Proceedings of 11th International Conference on Stability of Ships and Ocean
Vehicles (STAB 2012), Athens, Greece, pp. 323–336.
Spyrou, K.J., Belenky, V., Themelis, N., Weems, K.M., 2014 “Detection of Surf-riding Behavior
of Ships in Irregular Seas,” Nonlinear Dynamics, Vol. 78, No. 1, pp 649–667.
Spyrou, K.J., Kontolefas, I., Themelis, N., 2016, “Dynamics of the Surf-Riding Behavior of A Ship
in A Multi-Chromatic Sea Environment” Proc 31st Symp. Naval Hydrodynamics, Monterey,
California, USA, 12 p.
Spyrou, K.J., Themelis, N., Kontolefas, I., 2018, “Nonlinear surge motions of a ship in bi-chromatic
following waves”, Communications in Nonlinear Science and Numerical Simulation, Vol. 56,
pp. 296–313.