4
$\begingroup$

The 3D stream function ${\bf \Psi}$ for a steady flow field can be defined as: $\rho{\bf u}={\bf \nabla}\times {\bf \Psi}$.

Where, ${\bf u}$ = velocity, $\rho$ = density.

Now, this ${\bf \Psi}$ can be in turn represented as: ${\bf \Psi}=\chi{\bf \nabla}\psi $. Where, $\chi$ and $\psi$ are stream surfaces. What is the physical significance of $\chi$ and $\psi$?

$\endgroup$
  • $\begingroup$ The author has directly defined the stream function as ψ = χ grad ψ. I am having trouble understanding why he has defined stream it in this manner. Specifically, I cannot figure out intuition behind that definition? $\endgroup$ – suraj kulkarni Sep 2 '18 at 15:23
  • 2
    $\begingroup$ Which author? Which page? $\endgroup$ – Qmechanic Sep 2 '18 at 20:29
  • $\begingroup$ Are you comfortable with the 2D version of this, where ##chi## is a constant? Does that make sense to you from the standpoint of physical interpretation? $\endgroup$ – Chet Miller Sep 2 '18 at 22:06
  • $\begingroup$ You should mention the book's name. As far as I know, if $\psi$ is the scalar stream-function for 2d or axisymmetric flow, then velocity field is given by the curl of the vector $[0,0,\psi]$. $\endgroup$ – Deep Sep 3 '18 at 6:02
3
$\begingroup$
  1. The 3D stream function ${\bf \Psi}$ exists such that $\rho{\bf u}={\bf \nabla}\times {\bf \Psi}$ in a simply connected region for an incompressible flow ${\bf \nabla}\cdot(\rho{\bf u})=0$ due to Poincare Lemma.

  2. Any 3D vector field ${\bf \Psi}={\bf \nabla}\varphi + \chi{\bf \nabla}\psi $ can be represented by 3 scalar Clebsch potentials.

  3. We can remove the $\varphi$ potential due to the gauge symmetry ${\bf \Psi}\to{\bf \Psi}+{\bf \nabla}\varphi$.

  4. The flow $\rho{\bf u}={\bf \nabla}\chi \times {\bf \nabla}\psi$ is along the 1D intersection of the two 2D equipotential surfaces for $\chi$ and $\psi$.

  5. $(\chi,\psi)$ is a canonical pair: The flow is invariant under canonical transformations $(\chi,\psi)\to (\chi^{\prime},\psi^{\prime})$.

$\endgroup$
  • $\begingroup$ +1 Great answer. I just realized that a divergence-free vector field allows it to be written as curl of another vector field. $\endgroup$ – Deep Sep 4 '18 at 5:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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