Source: Physical Review E, Vol. 70, No. 6, pp.

066306, 2004;
DOI: 10.1103/PhysRevE.70.066306

Slamming of a Breaking Wave on a Wall

Jian-Jun Shu
School of Mechanical & Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798

ABSTRACT during the small-time impact process. The essential mechanism

involved in the impact process can be described by the theoretical
This paper is intended to study impact forces of breaking waves treatment of potential flow. A small portion of the breaker tip is
on a rigid wall based on a nonlinear potential-flow theory. This is initially cut off to produce a finite wetted area on the wall and a high
a model problem for some technologically important design issues spike in the consequent impact results from an acceleration of water
such as the impact of breaking waves on ships, coastal and offshore towards the wall. We are interested here in the short-time successive
structures. We are interested in the short-time successive triggering of nonlinear effects using a small-time expansion of the full,
triggering of nonlinear effects using a small-time expansion. The nonlinear initial/boundary value problem. The leading small-time
analytical solutions for the impact force on a rigid wall and the expansion is taken to include the accelerating effect. The analytical
free-surface profile are derived. solutions for the hydrodynamic force on a wall and the free-surface
profile are derived. It is worth to mention to this end that the technique
KEY WORDS: Breaking wave, plunging wave, impulsive pressure proposed by Chwang (1978) and extended by Chwang (1983), Liu
(1986) and King & Needham (1994) to investigate the earthquake
INTRODUCTION effect on dams has been adopted here in the mathematical treatment to
the present problem although physical settings are different.
There is a long history of experimental and theoretical studies to
determine impact forces acting on a rigid wall, which is suddenly GOVERNING EQUATIONS
started from rest and made to move towards a fluid taper. The problem
is motivated by the impact of breaking waves on ships, coastal and We consider a rigid horizontal wall, being suddenly started from rest
offshore structures, which is one of the most severe environmental and made to move vertically with constant acceleration a0 towards a
loads on structures. The impact due to a breaking wave striking a wall two-dimensional fluid taper with semi-angle  (0    1/ 2) . A
is of high intensity and short duration. This is attributed to the direct
collision between a fluid and a wall surface. The direct collision of a definition sketch of the flow is shown in Figure 1. The axis of the fluid
breaking wave with a wall generates an impulsive pressure on the wall. taper is perpendicular to the wall. Let us nondimensionalize time t by
This is similar to the problem of initial-stage water impact. ( L / a0 )1 / 2 , distance ( x, y ) by L , velocity (u, v) by (a0 L)1 / 2 , pressure
Unfortunately, existing wave theories based on small- and finite- p by a0 L , where L is the wetted wall semi-length when the
amplitude assumptions cannot be directly adopted to evaluate the
breaking wave force on a wall due to the highly nonlinear and transient breaking wave just touches the wall at time t  0 and  is the density
nature of the problem. of the fluid. A mathematical statement of the above problem can now
In reviewing the previous studies, one of the most important and be written as
unresolved questions is how the initial stage of the breaking wave
impingement on the wall can be properly characterized and simulated. u v
  0, (1)
Cumberbatch (1960) considered the case of symmetric normal impact x y
of a water wedge on a wall and Zhang et al. (1996) extended his work
u u u p
to an oblique impact. These two works stemmed from an ad hoc u v  , (2)
assumption on the free-surface profiles close to the wall: in t x y x
Cumberbatch (1960), a linear function was assumed, while in Zhang et v v v p
al. (1996), an exponential function was used. u v  . (3)
t x y y
In the present study the free-surface profiles are analytically
determined without prescribed functions. Effects of gravity, viscosity
and surface tension can be neglected since inertia forces are dominant
For negative time t  0 everything is at rest,
MATHEMATICAL ANALYSIS
u  v  0 ,   0 for t  0 , (4)
The full nonlinear initial/boundary value problem consists of
equations (1)-(3) with conditions (4)-(8). These equations are solved
where  is the free surface ``elevation'' in the x direction beyond the
analytically by employing a small-time expansion. We assume that
undisturbed surface. On the surface, the kinematic and dynamic
boundary conditions require
u( x , y , t )  u1 ( x , y)t  O(t 2 ) , v( x , y , t )  v1 ( x , y)t  O(t 2 ) , (9)
 
u v , p  0 on x  1  ytan( )   ( y, t ) . (5) p( x , y , t )  p0 ( x , y)  O(t ) , ( y , t )  2 ( y)t 2  O(t 3 ) . (10)
t y

On the wall surface, the normal velocity of fluid particles must be the The leading-order equations are
same as that of the wall at all time
u1 v1 p p
  0 , u1   0 , v1   0 (11)
v  a0t on y  a0t / 2.
2
(6) x y x y

On the axis of the fluid taper, the normal velocity of the fluid must subject to the conditions
vanish from consideration of symmetry about the axis of the fluid
taper, u1  22 , p0  0 on x  1  ytan( ) , (12)

u  0 on x  0. (7) v1  a0 on y  0 , (13)

The pressure vanishes at infinity, u1  0 on x  0 , (14)

p  0 as y  . (8)
p0  0 as y  . (15)

y It is clear that pressure p0 satisfies the Laplace equation

z plane
A C  2 p0  2 p0
  0. (16)
x 2 y 2

L Introducing a complex-conjugate function q0 with respect to p0 , we

a0
(y,t) can construct an analytic function

f0 ( z)  p0  iq0 , z  x  iy. (17)

O x
B As shown in Figure 1, the conformal mapping

 ( )ei (1 / 2  )

w
z  1 1 / 2
(  1) 1 / 2 d (18)
 (1/ 2   ) 1

given by the Schwarz-Christoffel transformation, maps the upper half

w plane
of the w plane (w    i ) onto the region occupied by the fluid.
Here  is the Gamma function defined by


(w)    w 1e d .
0
A O B C
Function f 0 is also analytic in the transformed variable w . On the

0 1 free surface, which corresponds to   1 on the positive real axis, p0
vanishes. On the axis of the fluid taper, which corresponds to the
Figure 1. Physical z plane is conformally mapped onto the upper half negative real axis in the w plane, we have p0 / n  0 , which means
of the w plane. that q0 is a constant. Without loss of generality, we may assume

The solution domain for this set of equations (1)-(3) with conditions q0  0 for   0 along the negative real axis. On the wall surface,
(4)-(8) is unknown at this stage of the analysis but is conveniently which corresponds to the line segment 0    1 , we take
described as p0 / n  a0 . Therefore, along the real axis in the w plane, we have

{( x , y , t ) :0  x  1  ytan( )  ( y, t ) , a0t 2 / 2  y   , 0  t  }.

Im( f0 )  0 on      0 , (19) where  denotes the Cauchy principal value. It is better to express
the above formula in the form that is suitable for easy computation.
 f  Differentiating (32) with respect to  , we obtain
Re 0   a0 on 0    1, (20)
 n 
dP0 a0(1   ) 1
1 / 2
(1   )
Re( f0 )  0 on 1    . (21)  3/ 2
d  (1   ) (1 / 2   )
1/ 2 0  
d
(33)
on 0    1.
If s( ) measures the distance from point B in Figure 1 to any point on
the wall surface, the Cauchy-Riemann equations give The integral on the right-hand side of (33) can be obtained by contour
integration. Thus
Im( f0 )  0 on      0 , (22)
dP0 a (1   )cot( )   (1   )  (1   ) 
Im( f0 )  a0 s( ) on 0    1, (23)  3 / 20    1/ 2 d 
d  (1   ) (1 / 2   )  
1/ 2 1/ 2 0  (   )  (34)

Re( f0 )  0 on 1     , (24) on 0    1.

Integrating (34), we have

where the distance s( ) is given by (18) as
8a0(1   )cot( )
( ) P0 ( ) 

1
s( )  1 / 2
(1   )
 1 / 2
d on 0    1. (25)  (1   )1/ 2 (1 / 2   )
3/ 2

 (1/ 2   )  (35)
 /2  1/ 2 d
  a0 s( ) on 0    1.
0 ( tan    )
2 1 / 2 
sin(2 )(tan ) 2
If we introduce a new analytic function g0 ( w) by
From boundary conditions (12) to (15), we have

g0 (w)  (1  w) 1 / 2
f 0 (w) , (26)
Im( f 0 / w)
 22 ( ) on   1. (36)
Im( z / w)   0
the boundary conditions for g0 ( w) are unmixed

After some mathematical manipulation, we obtain

Im( g0 )  0 on      0 , (27)

a0 1 / 2 (1/ 2   )(1   ) B(1/ 2   ;   ;1/  )

Im( g0 )  a0 (1   )1 / 2 s( ) on 0    1, (28) 2 ( )  on   1. (37)
2(  1)sin[(1/ 2   ) ](3 / 2   )

Im( g0 )  0 on 1    . (29)
where B is the incomplete Beta function

The analytic function g0 ( w) which is regular in the upper half w 

B( ;  ; )     1 (1   )  1 d .
plane and vanishes at infinity can be obtained from the Schwarz 0

integral formula
Impact free surface profiles for different semi-angle tapers are
1 
Im( g0 ) shown in Figures 2. It has been found that the free-surface profile
g0 ( w)   d . (30) 2 ( y) close to the wall is proportional to y 2 /(1 2 ) , which is neither
    w
linear in Cumberbatch's assumption (1960) nor exponential in Zhang et
Substituting (26) -- (29) into (30), we have al.'s assumption (1996). Figure 3 shows that the impact hydrodynamic
pressure increases as  increases, whereas the pressure decreases as
a0 (1  w)1 / 2 s( )  increases. Therefore, the pressure distribution for   0 is the
 (1   )
1
f 0 ( w)  d . (31)
 0 1/ 2
(  w) maximum envelope of all pressure distributions. The maximum
pressure always occurs on the free surface. It is clear that negative
impact pressure appears near the axis of the fluid taper and consequent
The impact pressure on the wall is the real part of f 0 ( w) for cavitation will be generated in that region.
0    1 . Using (25) and integrating by parts, we have on 0    1

a0 (1   )
P0 ( )  Re( f 0   0 ) 
 3 / 2(1 / 2   )
(32)
(1   )1/ 2  (1   )1/ 2
   1/ 2 (1   ) 1/ 2 ln
1
d ,
0 (1   )1/ 2  (1   )1/ 2
 y Cumberbatch, E (1960). “The impact of a water wedge on a wall,”
Journal of Fluid Mechanics, Vol. 7, pp. 353-374.

King, AC, and Needham, DJ (1994). “The initial development of a jet

1.0 caused by fluid,body and free-surface interaction. Part 1. A uniformly
  1/ 6 accelerating plate,” Journal of Fluid Mechanics, Vol. 268, pp. 89-101.
  1/ 3 Liu, PL-F (1986). “Hydrodynamic pressures on rigid dams during
0.5
earthquakes,” Journal of Fluid Mechanics, Vol. 165, pp. 131-145.
x Zhang, S, Yue, DKP, and Tanizawa, K (1996). “Simulation of
plunging wave impact on a vertical wall,” Journal of Fluid Mechanics,
3
Vol. 327, pp. 221-254
0 1 2

Figure 2. Impact free surface shapes 2 ( ) / a0 for various semi-angle

 tapers.

  1/ 6

P0 ( ) / a 0
80

40

  1/ 3
20

x
0
0.2 0.4 0.6 0.8 1.0

-20

Figure 3. Impact pressure distributions P0 ( ) / a0 on the wall for

various semi-angle  tapers.

CONCLUSIONS

An analytical approach is pursued to study the impact force of a

breaking wave on a rigid wall. The initial stage of the impact is
characterized by an impact of a two-dimensional liquid taper acting on
the wall with a prescribed acceleration. The problems of the impact
forces of breaking waves impingement on the wall and the free-surface
profile have been solved analytically by using a small-time expansion.
Explicit analytical formulae for evaluating the impact pressure and the
free-surface profile have been given. It has been found that the free-
surface profile close to the wall is neither linear in Cumberbatch's
assumption (1960) nor exponential in Zhang et al.'s assumption (1996).

REFERENCES

Chwang, AT (1978). “Hydrodynamic pressures on sloping dams

during earthquakes. Part 2. Exact theory,” Journal of Fluid Mechanics,
Vol. 87, pp. 343-348.

Chwang, AT (1983). “Nonlinear hydrodynamic pressure on an

accelerating plate,” The Physics of Fluids A, Vol. 26, pp. 383-387.