Waves in a pool

Take 15 minutes to prepare this exercise.

Then, if you lack ideas to begin, look at the given clue and start searching for the solution.

A detailed solution is then proposed to you.

If you have more questions, feel free to ask them on the forum.

Consider a perfect incompressible fluid (density \(\rho\)) in a basin of width \(L\) (along Oy).

A wave propagates along (Ox). The velocity field is of the form :

\(\vec v=v_x(x,z,t)\;\vec u_x+v_z(x,z,t)\;\vec u_z\)

The pressure \(P(x,z,t)\) is hydrostatic in the canal.

Note \(H+\xi (x,t)\) the height of the water in the canal (\(H\) is a constant).

The flow is assumed irrotational.

Question

Applying the Euler equation to a fluid particle, find a differential equation between \(v_x\) and \(\xi\) assuming small movements.

Show that \(v_x\) is indépendent of \(z\).

Solution

The Euler equation is :

\(\rho \left( {\frac{{\partial \vec v}}{{\partial t}} + (\vec v.\overrightarrow {grad} )\vec v} \right) = - \overrightarrow {grad} P + \rho \vec g\)

The pressure is hydrostatic, which means it is given by the expression obtained in fluid statics, namely :

\(P(x,z,t) = {P_0} + \rho g(H + \xi (x,t) - z)\)

Assuming small movements, it keeps only the first order terms (similar approximation to the acoustic approximation for the study of waves in fluids).

The Euler equation becomes :

\(\rho \frac{{\partial \vec v}}{{\partial t}} = - \overrightarrow {grad} P + \rho \vec g\)

In projection along the horizontal, the differential equation is obtained between \(v_x\) and \(\xi\) :

\(\rho \frac{{\partial {v_x}}}{{\partial t}} = - \frac{{\partial P}}{{\partial x}} = - \rho g\frac{{\partial \xi }}{{\partial x}}\)

Is :

\( \frac{{\partial {v_x}}}{{\partial t}} = -  g\frac{{\partial \xi }}{{\partial x}}\)

The flow is incompressible, so :

\(div(\vec v) = \frac{{\partial {v_x}}}{{\partial x}} + \frac{{\partial {v_z}}}{{\partial z}} = 0\)

In order of magnitude, if we note \(\lambda\) the wavelength of the wave (in the direction (Ox)) and \(a\) a characteristic dimension of the vertical movement (with \(a<<\lambda\)) :

\(\frac{{\left| {{v_x}} \right|}}{\lambda } \approx \frac{{\left| {{v_z}} \right|}}{a}\;\;\;\;so\;\;\;\;\;\left| {{v_z}} \right| \approx \frac{a}{\lambda }\left| {{v_x}} \right| < < \left| {{v_x}} \right|\)

We can therefore neglect the vertical component of velocity.

The flow is irrotational :

\(\overrightarrow {rot}( \vec v) = \vec 0\)

This leads to, neglecting \(v_z\) :

\(\frac{{\partial {v_x}}}{{\partial z}} = 0\)

Therefore, \(v_x\) does not depend on \(z\).

Question

Compute the volume flow rate \(Q_v\).

Make a mass balance on a slice of thickness \(dx\).

Deduce a differential equation between \(v_x\) and \(\xi\) assuming small movements.

Solution

The volume flow rate is :

\({Q_v}(x,t) = {v_x}(x,t)L(H + \xi (x,t))\)

Mass balance on a slice of thickness \(dx\), located between \(x\) and \(x+dx\) :

\(dm\) is the mass variation in this slice. It can be written in two ways :

\(dm = \rho L(H + \xi (x,t + dt))dx - \rho L(H + \xi (x,t))dx = \rho L\frac{{\partial \xi (x,t)}}{{\partial t}}dxdt\)

Or :

\(dm = \rho {Q_v}(x,t)dt - \rho {Q_v}(x + dx,t)dt = - \rho \frac{{\partial {Q_v}}}{{\partial x}}dxdt = - \rho L\frac{\partial }{{\partial x}}\left( {{v_x}(x,t)(H + \xi (x,t))} \right)dxdt\)

For identification :

\(\frac{{\partial \xi (x,t)}}{{\partial t}} = - \frac{\partial }{{\partial x}}\left( {{v_x}(x,t)(H + \xi (x,t))} \right)\)

The first order :

\(\frac{{\partial \xi (x,t)}}{{\partial t}} = - H\frac{{\partial {v_x}(x,t)}}{{\partial x}}\)

Question

Show that \(v_x\) and \(\xi\) check each classical wave equation.

Deduce the velocity of propagation of waves in the channel.

Solution

It was also the two equations :

\(\frac{{\partial \xi (x,t)}}{{\partial t}} = - H\frac{{\partial {v_x}(x,t)}}{{\partial t}}\)

And :

\( \frac{{\partial {v_x}(x,t)}}{{\partial t}} = -  g\frac{{\partial \xi(x,t) }}{{\partial x}}\)

By decoupling these two equations, we arrive at the d'Alembert equation :

\(\frac{{{\partial ^2}\xi (x,t)}}{{\partial {x^2}}} - \frac{1}{{{c^2}}}\frac{{{\partial ^2}\xi (x,t)}}{{\partial {t^2}}} = 0\;\;\;\;\;and\;\;\;\;\;\frac{{{\partial ^2}{v_x}(x,t)}}{{\partial {x^2}}} - \frac{1}{{{c^2}}}\frac{{{\partial ^2}{v_x}(x,t)}}{{\partial {t^2}}} = 0\)

Where the wave propagation speed is :

\(c=\sqrt{gL}\)

SimulationA video about standing waves
Making standing waves

Pour lire la vidéo, cliquer ici :