Difference between revisions of "Long Wavelength Approximations"

From WikiWaves
Jump to navigationJump to search
 
(3 intermediate revisions by one other user not shown)
Line 2: Line 2:
 
  | chapter title = Long Wavelength Approximations
 
  | chapter title = Long Wavelength Approximations
 
  | next chapter = [[Wave Scattering By A Vertical Circular Cylinder]]
 
  | next chapter = [[Wave Scattering By A Vertical Circular Cylinder]]
  | previous chapter = [[Added-Mass, Damping Coefficients And Exciting Forces]]  
+
  | previous chapter = [[Linear Wave-Body Interaction]]  
 
}}
 
}}
  
Line 74: Line 74:
 
<center><math> \forall =\pi a^2, \quad A_{11} = \pi \rho a^2 \,</math></center>
 
<center><math> \forall =\pi a^2, \quad A_{11} = \pi \rho a^2 \,</math></center>
  
<center><math> \frac{\partial u}{\partial t} = \mathrm{Re} \left\{ i\omega^2 e^{-K h + i \omega t} \right\} </math></center>
+
<center><math> \frac{\partial u}{\partial t} = \mathrm{Re} \left\{ i\omega^2 e^{-kh + i \omega t} \right\} </math></center>
  
 
Thus:
 
Thus:
Line 88: Line 88:
 
Here we will evaluate <math> \frac{\partial u}{\partial t} \,</math> on the axis of the platform and use a strip wise integration to evaluate the total hydrodynamic force.
 
Here we will evaluate <math> \frac{\partial u}{\partial t} \,</math> on the axis of the platform and use a strip wise integration to evaluate the total hydrodynamic force.
  
<center><math> u = \frac{\partial \Phi}{\partial x} = \mathrm{Re} \left\{ \frac{i g A}{\omega} (- i K) e^{Kz-i K x + i\omega t} \right\} </math></center>
+
<center><math> u = \frac{\partial \Phi}{\partial x} = \mathrm{Re} \left\{ \frac{i g A}{\omega} (- i k) e^{kz-i k x + i\omega t} \right\} </math></center>
  
<center><math> = \mathrm{Re} \left\{ \omega A e^{Kz+i\omega t} \right\}_{x=0} \,</math></center>
+
<center><math> = \mathrm{Re} \left\{ \omega A e^{kz+i\omega t} \right\}_{x=0} \,</math></center>
  
<center><math> \frac{\partial u}{\partial t} (z) = \mathrm{Re} \left\{ \omega A ( i \omega) e^{Kz+i\omega t} \right\} </math></center>
+
<center><math> \frac{\partial u}{\partial t} (z) = \mathrm{Re} \left\{ \omega A ( i \omega) e^{kz+i\omega t} \right\} </math></center>
  
<center><math> = - \omega^2 A e^{Kz} \sin \omega t \,</math></center>
+
<center><math> = - \omega^2 A e^{kz} \sin \omega t \,</math></center>
  
 
The differential horizontal force over a strip <math> \mathrm{d} z \,</math> at a depth <math> z \,</math> becomes:
 
The differential horizontal force over a strip <math> \mathrm{d} z \,</math> at a depth <math> z \,</math> becomes:
Line 102: Line 102:
 
<center><math> \rho ( \pi a^2 + \pi a^2 ) \frac{\partial u}{\partial t} \mathrm{d} z \,</math></center>
 
<center><math> \rho ( \pi a^2 + \pi a^2 ) \frac{\partial u}{\partial t} \mathrm{d} z \,</math></center>
  
<center><math> 2 \pi \rho a^2 \left( - \omega^2 A e^{Kz} \right) \sin \omega t \mathrm{d} z </math></center>
+
<center><math> 2 \pi \rho a^2 \left( - \omega^2 A e^{kz} \right) \sin \omega t \mathrm{d} z </math></center>
  
 
The total horizontal force over a truncated cylinder of draft <math>T\,</math> becomes:
 
The total horizontal force over a truncated cylinder of draft <math>T\,</math> becomes:
  
<center><math> F_x = \int_{-T}^{0} \mathrm{d}Z \mathrm{d}F = -2\pi\rho a^2 \omega^2 A \sin \omega t \int_{-T}^0 e^{Kz} \mathrm{d}z </math></center>
+
<center><math> F_x = \int_{-T}^{0} \mathrm{d}Z \mathrm{d}F = -2\pi\rho a^2 \omega^2 A \sin \omega t \int_{-T}^0 e^{kz} \mathrm{d}z </math></center>
  
<center><math> X_1 \equiv F_x = - 2 \pi \rho a^2 \omega^2 A \sin \omega t \cdot \frac{1-e^{-KT}}{K} </math></center>
+
<center><math> X_1 \equiv F_x = - 2 \pi \rho a^2 \omega^2 A \sin \omega t \cdot \frac{1-e^{-kT}}{k} </math></center>
  
This is a very useful and practical result. It provides an estimate of the surge exciting force on one leg of a possibly multi-leg platform as <math> T \to \infty; \quad \frac{1-e^{-KT}}{K} \to \frac{1}{K}\,</math>
+
This is a very useful and practical result. It provides an estimate of the surge exciting force on one leg of a possibly multi-leg platform as <math> T \to \infty; \quad \frac{1-e^{-kT}}{k} \to \frac{1}{k}\,</math>
  
 
=== Horizontal force on multiple vertical cylinders in any arrangement ===
 
=== Horizontal force on multiple vertical cylinders in any arrangement ===
Line 116: Line 116:
 
The proof is essentially based on a phasing argument. Relative to the reference frame,
 
The proof is essentially based on a phasing argument. Relative to the reference frame,
  
<center><math> \Phi_I = \mathrm{Re} \left\{ \frac{i g A}{\omega} e^{Kz-iKx + i\omega t} \right\} \,</math></center>
+
<center><math> \Phi_I = \mathrm{Re} \left\{ \frac{i g A}{\omega} e^{kz-ikx + i\omega t} \right\} \,</math></center>
  
 
Expressing the incident wave relative to the local frames by introducing the phase factors,
 
Expressing the incident wave relative to the local frames by introducing the phase factors,
  
<center><math> \mathbf{P}_i = e^{-iKx_i} </math></center>
+
<center><math> \mathbf{P}_i = e^{-ikx_i} </math></center>
  
 
and letting
 
and letting
Line 128: Line 128:
 
Then relative to the i-th leg,
 
Then relative to the i-th leg,
  
<center><math> \Phi_I^{(i)} = \mathrm{Re} \left\{ \frac{ i g A}{\omega} e^{Kz - iK\xi_i + i\omega t} \mathbf{P}_i \right\} \quad i=1,\cdots,N </math></center>
+
<center><math> \Phi_I^{(i)} = \mathrm{Re} \left\{ \frac{ i g A}{\omega} e^{kz - ik\xi_i + i\omega t} \mathbf{P}_i \right\} \quad i=1,\cdots,N </math></center>
  
 
Ignoring interactions between legs, which is a good approximation in long waves, the total exciting force on an n-cylinder platform is
 
Ignoring interactions between legs, which is a good approximation in long waves, the total exciting force on an n-cylinder platform is
Line 140: Line 140:
 
=== Surge exciting force on a 2D section ===
 
=== Surge exciting force on a 2D section ===
  
<center><math> \Phi_I = \mathrm{Re} \left\{ \frac{ i g A}{\omega} e^{Kz-iKx+i\omega t} \right\} \,</math></center>
+
<center><math> \Phi_I = \mathrm{Re} \left\{ \frac{ i g A}{\omega} e^{kz-ikx+i\omega t} \right\} \,</math></center>
  
<center><math> u=\mathrm{Re} \left\{ \frac{ i g A}{\omega} (- i K ) e^{Kz-iKx+i\omega t} \right\} \,</math></center>
+
<center><math> u=\mathrm{Re} \left\{ \frac{ i g A}{\omega} (- i k ) e^{kz-ikx+i\omega t} \right\} \,</math></center>
  
 
<center><math> \frac{\partial u}{\partial t} = \mathrm{Re} \left\{ \frac{ i g A}{\omega} \left(- i \frac{\omega^2}{g} \right) (i\omega) e^{i\omega t} \right\}_{x=0, z=0} \,</math></center>
 
<center><math> \frac{\partial u}{\partial t} = \mathrm{Re} \left\{ \frac{ i g A}{\omega} \left(- i \frac{\omega^2}{g} \right) (i\omega) e^{i\omega t} \right\}_{x=0, z=0} \,</math></center>
Line 164: Line 164:
 
where <math>A_w\,</math> is the body water plane area in 2D or 3D. <math>A\,</math> is the wave amplitude. This can be shown to be the leading order contribution from the Froude-Krylov force:
 
where <math>A_w\,</math> is the body water plane area in 2D or 3D. <math>A\,</math> is the wave amplitude. This can be shown to be the leading order contribution from the Froude-Krylov force:
  
<center><math> \mathbf{X}_3^{FK} = \rho g A \iint_{S_B} e^{Kz-iKx} n_3 \mathrm{d}S \,</math></center>
+
<center><math> \mathbf{X}_3^{FK} = \rho g A \iint_{S_B} e^{kz-ikx} n_3 \mathrm{d}S \,</math></center>
  
 
Using the Taylor series expansion,
 
Using the Taylor series expansion,
  
<center><math> e^{Kz-iKx} = 1 + ( Kz - iKx ) + O ( KB )^2 \,</math></center>
+
<center><math> e^{kz-ikx} = 1 + ( kz - ikx ) + O ( kB )^2 \,</math></center>
  
 
It is easy to verify that <math>\mathbf{X}_3 \to \rho g A A_w \,</math>.
 
It is easy to verify that <math>\mathbf{X}_3 \to \rho g A A_w \,</math>.
  
The scattering contribution is of order <math> KB\,</math>. For submerged bodies, <math> \mathbf{X}_3^{FK}=O(KB)\,</math>.
+
The scattering contribution is of order <math> kB\,</math>. For submerged bodies, <math> \mathbf{X}_3^{FK}=O(kB)\,</math>.
  
 
-----
 
-----

Latest revision as of 01:55, 12 February 2010

Wave and Wave Body Interactions
Current Chapter Long Wavelength Approximations
Next Chapter Wave Scattering By A Vertical Circular Cylinder
Previous Chapter Linear Wave-Body Interaction



Introduction

Very frequently the length of ambient waves [math]\displaystyle{ \lambda \, }[/math] is large compared to the dimension of floating bodies. For example the length of a wave with period [math]\displaystyle{ T=10 \mbox{s}\, }[/math] is [math]\displaystyle{ \lambda \simeq T^2 + \frac{T^2}{2} \simeq 150\mbox{m} \, }[/math]. The beam of a ship with length [math]\displaystyle{ L=100\mbox{m}\, }[/math] can be [math]\displaystyle{ 20\mbox{m}\, }[/math] as is the case for the diameter of the leg of an offshore platform.

GI Taylor's formula

Consider a flow field given by

[math]\displaystyle{ U(x,t):\ \mbox{Velocity of ambient unidirectional flow} \, }[/math]

[math]\displaystyle{ P(x,t):\ \mbox{Pressure corresponding to} \ U(x,t) \, }[/math]

[math]\displaystyle{ \lambda \sim \frac{|U|}{|\nabla U|} \gg B \ = \ \mbox{Body characteristic dimension} \, }[/math]

In the absence of viscous effects and to leading order for [math]\displaystyle{ \lambda \gg B \, }[/math]:

[math]\displaystyle{ F_x = - \left( \forall + \frac{A_{11}}{\rho} \right) \left. \frac{\partial P}{\partial x} \right|_{x=0} }[/math]

where

[math]\displaystyle{ \ F_x: \ \mbox{Force in x-direction} \, }[/math]
[math]\displaystyle{ \ \forall: \ \mbox{Body displacement}\, }[/math]
[math]\displaystyle{ \ A_{11}: \ \mbox{Surge added mass} \, }[/math]

Derivation using Euler's equations

An alternative form of GI Taylor's formula for a fixed body follows from Euler's equations:

[math]\displaystyle{ \frac{\partial U}{\partial t} + U \frac{\partial U}{\partial x} \simeq - \frac{1}{\rho} \frac{\partial P}{\partial x} }[/math]

Thus:

[math]\displaystyle{ F_x = \left( \rho \forall + A_{11} \right) + \left( \frac{\partial U}{\partial t} + U \frac{\partial U}{\partial x} \right)_{x=0} }[/math]

If the body is also translating in the x-direction with displacement [math]\displaystyle{ x_1(t)\, }[/math] then the total force becomes

[math]\displaystyle{ \ F_x = \left( \rho\forall+A_{11} \right) \left( \frac{\partial U}{\partial t} + U \frac{\partial U}{\partial x} \right) - A_{11} \frac{\mathrm{d}^2x_1(t)}{\mathrm{d}t^2} }[/math]

Often, when the ambient velocity [math]\displaystyle{ U\, }[/math] is arising from plane progressive waves, [math]\displaystyle{ \left| U \frac{\partial U}{\partial x} \right| = 0(A^2) \, }[/math] and is omitted. Note that [math]\displaystyle{ U\, }[/math] does not include disturbance effects due to the body.

Applications of GI Taylor's formula in wave-body interactions

Archimedean hydrostatics

[math]\displaystyle{ P=-\rho g z, \quad \frac{\partial P}{\partial z} = - \rho g \, }[/math]
[math]\displaystyle{ F_z = - ( \forall + \phi ) \frac{\partial P}{\partial z} = \rho g \forall }[/math]
[math]\displaystyle{ \phi: \ \mbox{no added mass since there is no flow} }[/math]

So Archimedes' formula is a special case of GI Taylor when there is no flow. This offers an intuitive meaning to the term that includes the body displacement.

Regular waves over a circle fixed under the free surface

[math]\displaystyle{ \Phi_I = \mathrm{Re} \left\{ \frac{i g A}{\omega} e^{kz-ikx+i\omega t} \right\}, \quad k=\frac{\omega^2}{g} \, }[/math]
[math]\displaystyle{ u=\frac{\partial \Phi_I}{\partial x} = \mathrm{Re} \left\{ \frac{i g A}{\omega} (-i k) e^{k z - i k x + i \omega t } \right \} }[/math]
[math]\displaystyle{ \mathrm{Re} \left\{ \omega A e^{ - k h +i \omega t} \right\}_{x=0,z=-h} }[/math]

So the horizontal force on the circle is:

[math]\displaystyle{ F_x = \left( \forall + \frac{A_{11}}{\rho} \right) \frac{\partial u}{\partial t} + O \left( z^2 \right) }[/math]
[math]\displaystyle{ \forall =\pi a^2, \quad A_{11} = \pi \rho a^2 \, }[/math]
[math]\displaystyle{ \frac{\partial u}{\partial t} = \mathrm{Re} \left\{ i\omega^2 e^{-kh + i \omega t} \right\} }[/math]

Thus:

[math]\displaystyle{ F_x = - 2 \pi a^2 \omega^2 A e^{-k h} \sin \omega t \, }[/math]

We can derive the vertical force along very similar lines. It is simply [math]\displaystyle{ 90^\circ\, }[/math] out of phase relative to [math]\displaystyle{ F_x\, }[/math] with the same modulus.

Horizontal force on a fixed circular cylinder of draft [math]\displaystyle{ T\, }[/math]

This case arises frequently in wave interactions with floating offshore platforms.

Here we will evaluate [math]\displaystyle{ \frac{\partial u}{\partial t} \, }[/math] on the axis of the platform and use a strip wise integration to evaluate the total hydrodynamic force.

[math]\displaystyle{ u = \frac{\partial \Phi}{\partial x} = \mathrm{Re} \left\{ \frac{i g A}{\omega} (- i k) e^{kz-i k x + i\omega t} \right\} }[/math]
[math]\displaystyle{ = \mathrm{Re} \left\{ \omega A e^{kz+i\omega t} \right\}_{x=0} \, }[/math]
[math]\displaystyle{ \frac{\partial u}{\partial t} (z) = \mathrm{Re} \left\{ \omega A ( i \omega) e^{kz+i\omega t} \right\} }[/math]
[math]\displaystyle{ = - \omega^2 A e^{kz} \sin \omega t \, }[/math]

The differential horizontal force over a strip [math]\displaystyle{ \mathrm{d} z \, }[/math] at a depth [math]\displaystyle{ z \, }[/math] becomes:

[math]\displaystyle{ \mathrm{d}F_z = \rho ( \forall + A_{11} ) \frac{\partial u}{\partial t} \mathrm{d} z \, }[/math]
[math]\displaystyle{ \rho ( \pi a^2 + \pi a^2 ) \frac{\partial u}{\partial t} \mathrm{d} z \, }[/math]
[math]\displaystyle{ 2 \pi \rho a^2 \left( - \omega^2 A e^{kz} \right) \sin \omega t \mathrm{d} z }[/math]

The total horizontal force over a truncated cylinder of draft [math]\displaystyle{ T\, }[/math] becomes:

[math]\displaystyle{ F_x = \int_{-T}^{0} \mathrm{d}Z \mathrm{d}F = -2\pi\rho a^2 \omega^2 A \sin \omega t \int_{-T}^0 e^{kz} \mathrm{d}z }[/math]
[math]\displaystyle{ X_1 \equiv F_x = - 2 \pi \rho a^2 \omega^2 A \sin \omega t \cdot \frac{1-e^{-kT}}{k} }[/math]

This is a very useful and practical result. It provides an estimate of the surge exciting force on one leg of a possibly multi-leg platform as [math]\displaystyle{ T \to \infty; \quad \frac{1-e^{-kT}}{k} \to \frac{1}{k}\, }[/math]

Horizontal force on multiple vertical cylinders in any arrangement

The proof is essentially based on a phasing argument. Relative to the reference frame,

[math]\displaystyle{ \Phi_I = \mathrm{Re} \left\{ \frac{i g A}{\omega} e^{kz-ikx + i\omega t} \right\} \, }[/math]

Expressing the incident wave relative to the local frames by introducing the phase factors,

[math]\displaystyle{ \mathbf{P}_i = e^{-ikx_i} }[/math]

and letting

[math]\displaystyle{ x = x_i + \xi_i \, }[/math]

Then relative to the i-th leg,

[math]\displaystyle{ \Phi_I^{(i)} = \mathrm{Re} \left\{ \frac{ i g A}{\omega} e^{kz - ik\xi_i + i\omega t} \mathbf{P}_i \right\} \quad i=1,\cdots,N }[/math]

Ignoring interactions between legs, which is a good approximation in long waves, the total exciting force on an n-cylinder platform is

[math]\displaystyle{ \mathbf{X}_1^N = \sum_{i=1}^N \mathbf{P}_i \mathbf{X}_1 \, }[/math]

The above expression gives the complex amplitude of the force with [math]\displaystyle{ \mathbf{X}_1\, }[/math] given in the single cylinder case.

The above technique may be easily extended to estimate the Sway force and Yaw moment on n-cylinders with little extra effort.

Surge exciting force on a 2D section

[math]\displaystyle{ \Phi_I = \mathrm{Re} \left\{ \frac{ i g A}{\omega} e^{kz-ikx+i\omega t} \right\} \, }[/math]
[math]\displaystyle{ u=\mathrm{Re} \left\{ \frac{ i g A}{\omega} (- i k ) e^{kz-ikx+i\omega t} \right\} \, }[/math]
[math]\displaystyle{ \frac{\partial u}{\partial t} = \mathrm{Re} \left\{ \frac{ i g A}{\omega} \left(- i \frac{\omega^2}{g} \right) (i\omega) e^{i\omega t} \right\}_{x=0, z=0} \, }[/math]
[math]\displaystyle{ = \mathrm{Re} \left\{ i \omega^2 A e^{i\omega t} \right\} = -\omega^2 A \sin \omega t \, }[/math]
[math]\displaystyle{ \mathbf{X}_1 = \left( \rho \forall + A_{11} \right) \frac{\partial u}{\partial t} = - \omega^2 A \sin \omega t ( \rho \forall + A_{11} ) \, }[/math]

If the body section is a circle with radius [math]\displaystyle{ a\, }[/math],

[math]\displaystyle{ \rho \forall = A_{11} = \pi\rho \frac{a^2}{2} \, }[/math]

So in long waves, the surge exciting force is equally divided between the Froude-Krylov and the diffraction components. This is not the case for Heave!

Heave exciting force on a surface piercing section

In long waves, the leading order effect in the exciting force is the hydrostatic contribution

[math]\displaystyle{ \mathbf{X}_i \sim \rho g A_w A \, }[/math]

where [math]\displaystyle{ A_w\, }[/math] is the body water plane area in 2D or 3D. [math]\displaystyle{ A\, }[/math] is the wave amplitude. This can be shown to be the leading order contribution from the Froude-Krylov force:

[math]\displaystyle{ \mathbf{X}_3^{FK} = \rho g A \iint_{S_B} e^{kz-ikx} n_3 \mathrm{d}S \, }[/math]

Using the Taylor series expansion,

[math]\displaystyle{ e^{kz-ikx} = 1 + ( kz - ikx ) + O ( kB )^2 \, }[/math]

It is easy to verify that [math]\displaystyle{ \mathbf{X}_3 \to \rho g A A_w \, }[/math].

The scattering contribution is of order [math]\displaystyle{ kB\, }[/math]. For submerged bodies, [math]\displaystyle{ \mathbf{X}_3^{FK}=O(kB)\, }[/math].


This article is based on the MIT open course notes and the original article can be found here

Ocean Wave Interaction with Ships and Offshore Energy Systems