SOLUTION OF LINEARIZED

BOUNDARY VALUE
PROBLEM
Bilge TUTAK
November 03, 2020

1 / 20
SOLUTION OF LINEARIZED
BOUNDARY VALUE
PROBLEM

2 / 20
 BOUNDARY VALUE DOMAIN
We have completed the linearization of the equations and boundary conditions. Now
we can tackle the solution of the Linearized Boundary Value Problem

Final form of the Linearized Boundary Value Problem

3 / 20
 BASIC EQUATION FOR SOLUTION
Laplace Equation ∇2 ϕ = 0 is a Linear Partial Di erential Equation. One method for
solving linear PDE’s is separation of variables. Assuming the solution can be
expressed as multiplication of functions;

ϕ(x, z, t) = X(x)Z(z)T (t)

If we substitute the above function in to the Laplace Equation,

2
∇ ϕ = Xxx ZT + X Zzz T = 0

If we divide the the equation with ϕ we will obtain

Xxx Zzz
+ = 0
X Z

4 / 20
The only way this can be possible is that both terms should be equal to a constant

Xxx Zzz
= − = Constant(Separation Constant)
X Z

Therefore the system turns in to two Ordinary Di erential Equations,

2
d X
− CX = 0
2
dx
2
d Z
+ CZ = 0
2
dz

5 / 20
CHARACTER OF SEPERATION CONSTANT C
For di erent characteristics of the Separation constant, the equation will yield a
di erent solution.

6 / 20
 1. C = k
2
,C > 0
2
d X
2 kx −kx
+ k X = 0 ⟹ X = Ae + Be (Growing solutions)
2
dx

2
d Z
2
− k Z = 0 ⟹ Z = D cos kz + E sin kz
2
dz

7 / 20
 2. C = 0 ,
2
d X
= 0 ⟹ X = Ax + B
2
dx

2
d Z
= 0 ⟹ Z = Dz + E
2
dz

Both solutions are growing without bounds, therefore they are no good.

8 / 20
 3. C = −k
2
,C < 0
2
d X
2
− k X = 0 ⟹ X = A cos kx + B sin kx
2
dx

2
d Z
2 kz −kz
+ k Z = 0 ⟹ Z = De + Ee
2
dz

9 / 20
 Out of these 3 solution sets, the only possible solution with the needed periodicity in
the x− direction would be the 3rd case. Therefore we can nd the “General Solution”
of the velocity potential,
kz −kz
ϕ(x, z, t) = (A cos kx + B sin kx) (De + Ee ) T (t)

We also need periodicity in time. Therefore a reasonable assumption for T (t) would
be a function similar to X(x). Then,
kz −kz
ϕ(x, z, t) = (A cos kx + B sin kx) (De + Ee ) (F cos σt + G sin σt)

10 / 20
 MANY SOLUTIONS
We want to apply the boundary conditions to solve for the unique solution of the
wave equation. However the above solution of the velocity potential is a little complex
to apply to the boundary conditions. Luckily Laplace Equation is a linear equation.
Therefore we can use the lineary of Laplace Equation and the Superposition Principle
to get simpler functions for velocity potential ϕ;
kz −kz
ϕ1 = A1 (De + Ee ) cos kx cos σt

kz −kz
ϕ2 = A2 (De + Ee ) sin kx sin σt

kz −kz
ϕ3 = A3 (De + Ee ) sin kx cos σt

kz −kz
ϕ4 = A4 (De + Ee ) cos kx sin σt
Assuming we use ϕ1

11 / 20
 BOTTOM BOUNDARY CONDITION
We have the bottom boundary condition as;

ϕz = 0 on z = −h

Di erentiating ϕ1 once with respect to z and applying the boundary condition,

kz −kz
ϕz = A1 (Dke − Eke ) cos kx cos σt|z=−h = 0

The only way the ϕz term can be zero is that the term with exponentials should be
equal to zero. Therefore,
−kh kh
(Dke − Eke ) = 0

12 / 20
 We can nd D in terms of E
2kh
D = Ee

If we substitute D and E in to the Z(z) equation we can nd the function,

kz −kz 2kh kz −kz

Z = De + Ee = Ee e + Ee

k(h+z) −k(h+z)
e + e
kh
= 2E e [ ]
2

cosh [k(h+z)]

Finally, ϕ1 becomes;

ϕ1 = A1 cosh [k(h + z)] cos kx cos σt

13 / 20
DYNAMIC FREE SURFACE BOUNDARY
CONDITION
−ϕt + gη = 0 on z = 0

If we leave η alone;

1 ∣
η = ϕt ∣
g ∣
z=0

η = −A1 σ cosh (kh) cos kx cos σt

Since at cos kx cos σt surface elevation is at max ηmax

H
= 1 =
2

14 / 20
H σ
= −A1 cosh (kh)
2 g

Then, we can nd A1 ,

Hg 1
A1 = −
2σ cosh (kh)

15 / 20
Finally we can obtain ϕ1 and all the other potential solutions ϕ2 , ϕ3 , ϕ4

H g cosh [k(h + z)]

ϕ1 = − cos kx cos σt
2σ cosh (kh)

H g cosh [k(h + z)]

ϕ2 = + sin kx sin σt
2σ cosh (kh)

H g cosh [k(h + z)]

ϕ3 = − sin kx cos σt
2σ cosh (kh)

H g cosh [k(h + z)]

ϕ4 = + cos kx sin σt
2σ cosh (kh)

16 / 20
 LINEAR COMBINATION OF SOLUTIONS
We want a linear combination of di erent ϕ′ s such that we want to obtain a
progressive wave.
Since ∇ ϕ = 0 is a linear PDE, any linear combination of solutions is also a solution.
2

Which one should we choose is determined by the other required criteria.

Progressive wave de nition

17 / 20
Consider a wave of permanent form traveling with Celerity in the positive direction,

L
C =
T

If we consider the speed in terms of certain amount of distance in certain time,

L x0 − 0 t0 x0
= ⟹ − = 0
T t0 − 0 T L

The above equation will hold if and only if η(0, 0) = η(x0 , t0 ) Using the above
information and generalizing, we can write a wave phase angle with 2π periodicity;

t x
θ = 2π ( − )
T L

18 / 20
 Finally making use of a simple trigonometric idendity:
sin (α − β) = sin α cos β − sin β cos α

Now if we make a linear combination of ϕ3 and ϕ4

H g cosh [k(h + z)]

ϕ = ϕ3 − ϕ4 = − (sin kx cos σt − cos kx sin σt)
2σ cosh (kh) 
=sin (kx−σt)

Finally, we can write the velocity potential of a wave as;

H g cosh [k(h + z)]

ϕ = − sin (kx − σt)
2σ cosh (kh)

Where we have Wave Number k and **Wave Angular Frequency σ

2π 2π
= =
L T

19 / 20
 SURFACE ELEVATION
Using DFSBC we can determine the surface elevation function η

1 ∣
η = ϕt ∣
g ∣
z=0

1 H g cosh [k(h + z)]

η = ⋅ [− cos (kx − σt)(−σ)]
g 2σ cosh (kh)
z=0

Then

H
η(x, t) = cos (kx − σt)
2

20 / 20