0
$\begingroup$

$\vec{w} = \begin{pmatrix}w_r \\ w_{phi} \end{pmatrix} = \begin{pmatrix}\frac{Q_0}{2 \pi r} \\ 0 \end{pmatrix} $

1) Show that the flow satisfys the continuity equation

2) Show that $\vec{w}$ has a potential $\Phi$

3) Calculate $\Phi$

Task 1

$\nabla \cdot \vec{w} = 0$ I can't derive $\frac{\partial w_r}{\partial r} + \frac{1}{r} \frac{\partial w_{\phi}}{\partial \phi} + \frac{\partial w_z}{\partial z} + \frac{w_r}{r} = 0$ from that. The last part $+ \frac{w_r}{r}$ remains a mystery to me.

Task 2

I would have used $\vec{\nabla} \times \vec{w} = 0$ However if I calculate that vector product I see that the equation is satisfyed, if I assume $w_z$ is zero since it is not given. From the solution however it states that: $w_z = \frac{w_{\phi}}{r} + \frac{\partial w_{\phi}}{\partial r} - \frac{1}{r}\frac{\partial w_r}{\partial \phi} = 0$ This is due to the fact that a plane flow can only have the third compoment $w_z$ different from zero. Neither can I imagine why that component is the only one that can be different from zeor nor where the equation comes from.

$\endgroup$
6
  • 1
    $\begingroup$ Welcome to Physics! Please note that Physics.StackExchange is not a homework help site. Please read this Meta post on asking homework-like questions and this Meta post for "check my work" problems. $\endgroup$
    – Kyle Kanos
    Commented Aug 5, 2015 at 14:55
  • $\begingroup$ @KyleKanos Thank you and I'm well aware of that. However I do not see why you linked those two posts here. I neither asked for a solution nor did I really calculate the solution. I had trouble understanding how the potential, curl etc. are linked to each other and I came to seek help for that specific concept. Please let me know what kind of changes you would like to see in my question and I will edit it accordingly. $\endgroup$
    – idkfa
    Commented Aug 5, 2015 at 15:23
  • $\begingroup$ Your question, as posted, is How do these two terms equal zero which is asking for solutions to a rather straight-forward differential. Outside of the cited question 2, I don't see anything about potentials here. If you are confused about potentials in fluid dynamics, ask about that & not how to calculate a few derivatives. $\endgroup$
    – Kyle Kanos
    Commented Aug 5, 2015 at 15:33
  • $\begingroup$ @KyleKanos Integrating in order to calculate the potential was not the problem but rather if the requirements to possess a potential are satisfied. Therefore I needed to know more about the derivatives. Would you propose changing the topic of the question for it to be in order? Or is this low quality with little chances of improvements that would actually help? I'll try to rephrase / reformat the question when I get home. $\endgroup$
    – idkfa
    Commented Aug 5, 2015 at 15:38
  • $\begingroup$ As I stated previously, if you are confused about potential flows in fluid dynamics, ask about what is confusing you and not how to calculate a few derivatives. $\endgroup$
    – Kyle Kanos
    Commented Aug 5, 2015 at 15:44

2 Answers 2

1
$\begingroup$

Task 1

Find a calculus textbook and look for cylindrical coordinates, you will find that the divergence of a vector in those coordinates is given by: $$\vec{\nabla}\cdot\vec{\omega} = \frac{1}{r}\frac{\partial}{\partial r}(r\omega_r) + \frac{1}{r}\frac{\partial}{\partial \phi}\omega_{\phi}+\frac{\partial\omega_z}{\partial z}$$ Now apply the product rule to the term $\frac{1}{r}\frac{\partial}{\partial r}(r\omega_r)$ and you will arrive at $\frac{\partial\omega_r}{\partial r} + \frac{\omega_r}{r}$. The additional mysterious term is to account for the curvature which is present compared to cartesian coordinates.

I find the unexpanded form generally easier to work with as in this case the $r$ dependence of $\omega_r$ cancels immediately and therefore the derivative is identically zero.

Task 2

The vector $\vec{\nabla}\times\vec{\omega}$ only has a $z$-component because the $\omega_z$-component is assumed zero. $$\vec{\nabla}\times\vec{\omega}=(\frac{1}{r}\frac{\partial}{\partial\phi}\omega_z-\frac{\partial}{\partial z}\omega_{\phi})\hat{e_r}-(\frac{1}{r}\frac{\partial}{\partial r}(r\omega_z)-\frac{\partial}{\partial z}\omega_r)\hat{e_\phi}+(\frac{1}{r}\frac{\partial}{\partial r}(r\omega_{\phi})-\frac{1}{r}\frac{\partial\omega_{r}}{\partial\phi})\hat{e_z}$$ since $\omega_{\phi}=\omega_{z}=0$, and $\omega_r$ is not a function of $z$; we have that only the $z$-component $\frac{1}{r}\frac{\partial}{\partial r}(r\omega_{\phi})-\frac{1}{r}\frac{\partial\omega_{r}}{\partial\phi}$ is non-zero. Expanding this as in Task 1 will give you the same form as in your solution and inserting the given $\omega_r$ and $\omega_{\phi}$ components will result in it being zero.

$\endgroup$
4
  • $\begingroup$ I don't understand what you mean by as in this case the r dependence of ωr cancels immediately. After calculating the derivative I see that the terms cancel each other out, I don't see how the dependence on $r$ cancels out. $\endgroup$
    – idkfa
    Commented Aug 5, 2015 at 12:45
  • $\begingroup$ Also when applying your equation for the two-dimensional vectors I don't end up with the equation used in the solution since there is no $z$ component. $\endgroup$
    – idkfa
    Commented Aug 5, 2015 at 12:47
  • $\begingroup$ $\omega_r=\frac{Q_0}{2\pi r}$ has an $r$ dependence, $r\omega_r=\frac{Q_0}{2\pi}$ has no $r$-dependece, therefore the derivative $\partial_r(r\omega_r)$ is zero. $\endgroup$
    – nluigi
    Commented Aug 5, 2015 at 12:48
  • $\begingroup$ Since you provided a two-dimensional $\vec{\omega}=i\omega_r+j\omega_{\phi}$ I disregarded the $z$-component altogether. However, if the $z$-component is present you should of course incorporate it $\endgroup$
    – nluigi
    Commented Aug 5, 2015 at 12:53
1
$\begingroup$

T1: $w_r/r$ is just $Q_o/2\pi r^2$ which cancels the first term. T2: If $w_z$ is not given, one should assume that the flow is perpendicular to the $z$-axis i.e. that $w_z=0$. In that case, the curl only has a $z$-component $$\frac{1}{\rho}(\frac{\partial(\rho w_\phi)}{\partial\rho}-\frac{\partial w_\rho}{\partial \phi}) $$ which turns out to be $0$.

$\endgroup$
3
  • $\begingroup$ I'm sorry could you elaborate on that a little bit more? I follow you that the flow is perpendicular to the z-axis. Hence the flow could only rotate around the z-axis, therefore the curl needs to have a z-component if so. Is that correct? Also I have never seen the equation you posted before. $\endgroup$
    – idkfa
    Commented Aug 5, 2015 at 12:53
  • 1
    $\begingroup$ A z-perpendicular flow can have both radial and rotational ($\phi$) components, but in this case it only has a radial component. For a flow without z-component and that is on top of that independent of the z-coordinate, the x and y components of the curl are automatically 0 (check curl formula in cylindrical coordinates). So, the z-component is the only one that we need to evaluate and even that one turns out to be 0. $\endgroup$
    – jac
    Commented Aug 5, 2015 at 12:58
  • $\begingroup$ Since without having a z-component there is no way to have r, $\phi$ components different from zero. I see now why only the z-component potentially can differ from zero. Thanks a lot. $\endgroup$
    – idkfa
    Commented Aug 5, 2015 at 13:04

Not the answer you're looking for? Browse other questions tagged or ask your own question.