Small Amplitude Water Wave Theory
Small Amplitude Water Wave Theory
3.1 Introduction
Assuming inviscid, incompressible fluid and irrotational flow, φ and Ψ exist, which
satisfy ∇ 2φ = ∇ 2 Ψ = 0 .
The kinematic boundary condition is a condition that describes the water particle
kinematics at a boundary (either fixed or moving). If we sit on a boundary (fixed or
moving) and move with the boundary, we do not feel any change of the surface that
constitutes the boundary. Mathematically, the rate of change of the surface must be zero,
i.e., the total derivative of the surface is zero on the surface. Let the surface of the
boundary be represented by F ( x, y , z , t ) = 0 . Then
DF ∂F ∂F ∂F ∂F
= +u +v +w = 0 on F =0
Dt ∂t ∂x ∂y ∂z
2
∂F ∂F ∂F ∂F ⎛ ∂F ⎞ ⎛ ∂F ⎞ ⎛ ∂F ⎞
2 2
− = u ⋅ ∇F = u ⋅ n ∇ F ; ∇F = i+ j+ k, ∇F = ⎜ ⎟ + ⎜⎜ ⎟⎟ + ⎜ ⎟
∂t ∂x ∂y ∂z ⎝ ∂x ⎠ ⎝ ∂y ⎠ ⎝ ∂z ⎠
− ∂F / ∂t
u⋅n= on F =0
∇F
3.3.2 BBC
z = − h ( x, y ) ⇒ F ( x, y , z , t ) = z + h ( x, y ) = 0
− ∂F / ∂t
u⋅n = = 0 on z = − h ( x, y )
∇F
Using
∂h ∂h
i+ j+k
∇F ∂x ∂y
n= =
∇F 2
⎛ ∂h ⎞ ⎛ ∂h ⎞
2
⎜ ⎟ + ⎜⎜ ⎟⎟ + 1
⎝ ∂x ⎠ ⎝ ∂y ⎠
we have
∂h ∂h
u +v +w
∂x ∂y
u⋅n = =0
2
⎛ ∂h ⎞ ⎛ ∂h ⎞
2
⎜ ⎟ + ⎜⎜ ⎟⎟ + 1
⎝ ∂x ⎠ ⎝ ∂y ⎠
∂h ∂h
w = −u −v on z = − h( x, y )
∂x ∂y
∂φ ∂φ ∂h ∂φ ∂h
− = + on z = − h( x, y )
∂z ∂x ∂x ∂y ∂y
On water surface,
z = η ( x, y , t ) ⇒ F ( x, y , z , t ) = z − η ( x, y , t ) = 0
− ∂F / ∂t ∂η / ∂t
u⋅n= = on z = η ( x, y , t )
∇F (∂η / ∂x )2 + (∂η / ∂y )2 + 1
Using
u = u i + v j + wk
∂η ∂η
− i− j+k
∇F ∂x ∂y
n= =
∇F 2
⎛ ∂η ⎞ ⎛ ∂η ⎞
2
⎜ ⎟ + ⎜⎜ ⎟⎟ + 1
⎝ ∂x ⎠ ⎝ ∂y ⎠
we have
∂η ∂η ∂η
−u −v +w=
∂x ∂y ∂t
∂η ∂η ∂η
w= +u +v on z = η ( x, y, t )
∂t ∂x ∂y
∂φ ∂η ∂φ ∂η ∂φ ∂η
− = − − on z = η ( x, y, t )
∂z ∂t ∂x ∂x ∂y ∂y
L ∼ O (h)
H / L = wave steepness << 1
η ∼ O (H )
t ∼ O (T )
x ∼ O (L)
z ∼ O (h) ∼ O(L)
∂φ ⎛H⎞
u= ∼ O⎜ ⎟
∂x ⎝T ⎠
⎛ HL ⎞
φ ∼ O⎜ ⎟
⎝ T ⎠
∂φ ∂η ∂φ ∂η ∂φ ∂η
− = − − on z = η ( x, y , t )
∂z ∂t ∂x ∂x ∂y ∂y
⎛H⎞ ⎛H⎞ ⎛H H⎞ ⎛H H⎞
⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟
⎝T ⎠ ⎝T ⎠ ⎝T L⎠ ⎝T L⎠
∂φ ∂η
∴− = on z = η ( x, y , t )
∂z ∂t
⎛ ∂φ ∂η ⎞ ⎛ ∂φ ∂η ⎞ ∂ ⎛ ∂φ ∂η ⎞ η 2 ∂ 2 ⎛ ∂φ ∂η ⎞
⎜ − − ⎟ = ⎜ − − ⎟ + η ⎜ − − ⎟ + ⎜− − ⎟ +L= 0
⎝ ∂z ∂t ⎠ z =η ⎝ ∂z ∂t ⎠ z = 0 ∂z ⎝ ∂z ∂t ⎠ z = 0 2 ∂z 2 ⎝ ∂z ∂t ⎠ z = 0
⎛H⎞ ⎛H⎞ ⎛H2 ⎞ ⎛H2 ⎞
⎜ ⎟ ⎜ ⎟ ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟
⎝T ⎠ ⎝T ⎠ ⎝ LT ⎠ ⎝ LT ⎠
∂φ ∂η
− = on z=0 (LKFSBC)
∂z ∂t
3.3.4 DFSBC
∂φ 1 2
− (
+ u + v 2 + w2 +
∂t 2
pη
ρ
)
+ gη = C (t ) on z = η ( x, y , t )
∂φ 1 2
−
∂t 2
( )
+ u + v 2 + w 2 + gη = C (t ) on z = η ( x, y , t )
⎛ HL ⎞ ⎛H2 ⎞
⎜ 2⎟ ⎜⎜ 2 ⎟⎟ (gH )
⎝T ⎠ ⎝T ⎠
⎛H⎞ ⎛ T2 ⎞
(1) ⎜ ⎟ << 1 ⎜⎜ g ⎟⎟ ~ O(1)
⎝L⎠ ⎝ L ⎠
∂φ
∴− + gη = C (t ) on z = η ( x, y , t )
∂t
⎛ ∂φ ⎞ ⎛ ∂φ ⎞ ∂ ⎛ ∂φ ⎞
⎜− + gη ⎟ = ⎜ − + gη ⎟ + η ⎜ − + gη ⎟ + L = C (t )
⎝ ∂t ⎠ z =η ⎝ ∂t ⎠ z =0 ∂z ⎝ ∂t ⎠ z =0
⎛ HL ⎞ ⎛H2 ⎛ H2 ⎞
⎞
⎜ 2 ⎟ (gH ) ⎜⎜ 2 ⎟⎟
⎜⎜ g ⎟⎟
⎝T ⎠ ⎝T ⎠
⎝ L ⎠
⎛ T2 ⎞ ⎛H⎞ ⎛ T H⎞
2
(1) ⎜⎜ g ⎟⎟ ⎜ ⎟ ⎜⎜ g ⎟⎟
⎝ L ⎠ ⎝L⎠ ⎝ L L⎠
∂φ
− + gη = C (t ) on z=0 (LDFSBC)
∂t
3.3.5 LBC
φ ( x , t ) = φ ( x + L, t )
φ ( x, t ) = φ ( x, t + T )
GE ∇ 2φ = 0 0 ≤ x ≤ L, − h ≤ z ≤ η
∂φ
BBC − = 0 on z = −h
∂z
∂φ ∂η
LKFSBC − = on z=0
∂z ∂t
∂φ
LDFSBC − + gη = C (t ) on z=0
∂t
φ ( x , t ) = φ ( x + L, t )
LBC
φ ( x, t ) = φ ( x, t + T )
φ ( x, z , t ) = Χ ( x ) Ζ( z )Τ(t )
Assume
2π
Τ(t ) = sin σt ; σ =
T
where σ = wave angular frequency. The above equation satisfies the periodicity
condition in time, that is
Now
φ ( x, z , t ) = Χ ( x ) Ζ( z ) sin σt
d 2Χ d 2Ζ
Ζ sin σt + Χ sin σt = 0
dx 2 dz 2
1 d 2Χ 1 d 2Ζ
+ =0
Χ dx 2 Ζ dz 2
The first term is a function of x alone, whereas the second term is a function of
z alone. Therefore, the two terms should be constants with opposite signs, that is,
1 d 2Χ
= −k 2
Χ dx 2
1 d 2Ζ
= k2
Ζ dz 2
The possible solutions that satisfy these ODE’s are given in Table 3.1 of textbook.
Among these, the solution which satisfies the periodicity condition in x is
2π
k=
L
Recalling that the superposition of solutions is valid for the Laplace equation, keep only
Using BBC,
∂φ
−
∂z
( )
= A cos kx kCe kz − kDe − kz sin σt = 0 on z = −h
Ce − kh − De kh = 0
C = De 2 kh
Therefore,
φ ( x, z, t ) = A cos kx(De 2 kh e kz + De − kz )sin σt
( )
= ADe kh cos kx e k ( h + z ) + e − k ( h + z ) sin σt
= G cos kx cosh k (h + z ) sin σt
Applying LDFSBC,
∂φ
− + gη = C (t ) on z = 0
∂t
− σG cos kx cosh kh cos σt + gη = C (t )
σG cosh kh C (t )
η= cos kx cos σt +
g g
H
η= cos kx cos σt
2
we have
gH
G=
2σ cosh kh
Finally,
gH cosh(h + z )
φ ( x, z , t ) = cos kx sin σt
2σ cosh kh
Applying LKFSBC,
∂φ ∂η
− = on z = 0
∂z ∂t
gkH sinh kh σH
− cos kx sin σt = − cos kx sin σt
2σ cosh kh 2
gkH σH
tanh kh =
2σ 2
∴ σ 2 = gk tanh kh (dispersion relationship)
The dispersion relationship gives relation among h , σ and k or the relation among
h , T and L . The dispersion relationship can be solved for k by Newton-Raphson
method for given σ and h . Approximate solutions are also available:
Eckart (1951):
σ 2h
σ 2 = gk tanh
g
Hunt (1979):
y σ 2h
(kh) 2 = y 2 + 6
; y= , d 1 ~ d 6 given in p. 72 of textbook
g
1+ ∑ dn y n
n =1
⎛ 2π ⎞ 2π
2
⎜ ⎟ =g tanh kh
⎝ T ⎠ L
gT 2
L= tanh kh
2π
gT 2
L0 =
2π
Then
L = L0 tanh kh
L L0
C= = tanh kh = C 0 tanh kh
T T
L gT
C0 = 0 =
T 2π
In summary,
Hg cosh(h + z )
φ ( x, z , t ) = cos kx sin σt
2σ cosh kh
H
η ( x, t ) = cos kx cos σt
2
Hg cosh(h + z )
φ ( x, z , t ) = − sin kx cos σt
2σ cosh kh
H
η ( x, t ) = sin kx sin σt
2
Hg cosh(h + z )
φ ( x, z , t ) = (cos kx sin σt − sin kx cos σt )
2σ cosh kh
H g cosh k (h + z )
=− sin(kx − σt )
2 σ cosh kh
H
η ( x, t ) = (cos kx cos σt + sin kx sin σt )
2
H
= cos(kx − σt )
2
Progressive wave propagating to positive x -direction:
sin( kx − σt ) ⇒ sin( kx + σt )
For the definitions of deepwater, shallow water, and intermediate depth water, see Fig.
3.12 of textbook. Also see Table 3.2 for asymptotic forms of hyperbolic functions.
σ 2 = gk tanh kh ≅ gk 2 h
⎛σ ⎞
2
C 2 = ⎜ ⎟ = gh
⎝k⎠
C = gh
The shallow water wave is non-dispersive, that is, all the waves with different
frequencies propagate at the same speed.
σ 2 = gk tanh kh ≅ gk
2π 2π 2πg gT 2
L0 = = 2 = =
k σ ⎛ 2π ⎞
2
2π
⎜ ⎟
g ⎝ T ⎠
L gT
C0 = 0 =
T 2π
3.5 Waves with Uniform Current U 0
∂φ
U0 = − ⇒ φ = −U 0 x
∂x
φ = −U 0 x + A cosh k (h + z ) cos(kx − σt )
∂φ
u=− = U 0 + kA cosh k (h + z ) sin( kx − σt )
∂x
DFSBC is
⎡ ∂φ 1 2
( ⎤
⎢− ∂t + 2 u + w + gη ⎥
2
)
⎣ ⎦ z =η
⎡ ∂φ 1 2 ∂ ⎡ ∂φ 1 2
= ⎢− ( ⎤
)
+ u + w 2 + gη ⎥ + η ⎢−
⎤
(
+ u + w 2 + gη ⎥ = C (t )
∂z ⎣ ∂t 2
)
⎣ ∂t 2 ⎦ z =0 ⎦ z =0
Note that w 2 is small compared to the linear terms but u 2 is not. Then DFSBC gives
− σA cosh kh sin( kx − σt ) +
1 2
2
( )
U 0 + 2kAU 0 cosh kh sin( kx − σt ) + gη = C (t )
or
U 02 Aσ ⎛ kU 0 ⎞
η ( x, t ) = − + ⎜1 − ⎟ cosh kh sin(kx − σt ) + C (t )
2g g ⎝ σ ⎠
Since
1 T U 02 1 T
η=
T ∫ 0
ηdt = 0 = −
2g T
+ ∫ 0
C (t )dt
we have
U 02
C (t ) = constant =
2g
Now
Aσ ⎛ kU 0 ⎞ H
η ( x, t ) = ⎜1 − ⎟ cosh kh sin(kx − σt ) = sin(kx − σt )
g ⎝ σ ⎠ 2
which gives
H g gH
A= =
2 ⎛ kU 0 ⎞ ⎛ U ⎞
σ ⎜1 − ⎟ cosh kh 2σ ⎜1 − 0 ⎟ cosh kh
⎝ σ ⎠ ⎝ C ⎠
Therefore,
gH cosh k (h + z )
φ ( x, z , t ) = −U 0 x + cos( kx − σt )
⎛ U 0 ⎞ cosh kh
2σ ⎜1 − ⎟
⎝ C ⎠
Applying KFSBC,
⎡ ∂η ∂φ ∂η ∂φ ⎤ ⎡ ∂η ∂φ ∂η ∂φ ⎤ ∂ ⎡ ∂η ∂φ ∂η ∂φ ⎤
⎢ ∂t − ∂x ∂x + ∂z ⎥ = ⎢ ∂t − ∂x ∂x + ∂z ⎥ + η ∂z ⎢ ∂t − ∂x ∂x + ∂z ⎥ + L = 0
⎣ ⎦ z =η ⎣ ⎦ z =0 ⎣ ⎦ z =0
∂η ∂η ∂φ
+U0 =− on z=0
∂t ∂x ∂z
H H kgH
−σ cos(kx − σt ) + U 0 k cos(kx − σt ) = − tanh kh cos(kx − σt )
2 2 ⎛ U0 ⎞
2σ ⎜1 − ⎟
⎝ C ⎠
gk
− σ + U 0k = − tanh kh
⎛ U0 ⎞
σ ⎜1 − ⎟
⎝ C ⎠
k gk
−1+U0 =− tanh kh
σ ⎛ U0 ⎞
σ ⎜1 − ⎟ 2
⎝ C ⎠
U gk
−1+ 0 = − tanh kh
C 2⎛ U0 ⎞
σ ⎜1 − ⎟
⎝ C ⎠
gk tanh kh
σ2 = 2
⎛ U0 ⎞
⎜1 − ⎟
⎝ C ⎠
2
⎛ U ⎞
σ ⎜1 − 0 ⎟ = gk tanh kh
2
⎝ C ⎠
⎛ U0 ⎞
σ ⎜1 − ⎟ = gk tanh kh
⎝ C ⎠
Finally,
σ = U 0 k + gk tanh kh
The second term on RHS indicates the angular frequency in no current, that is the
angular frequency in moving frame of reference, while the total indicates the angular
frequency in stationary frame of reference.