Modeling creeping (Stokes) flow past a solid sphere

Dear All

I recently started using FEniCS, and I want to set up and solve the classic fluid mechanics problem of creeping flow past a solid sphere, which has an analytical solution [http://web.mit.edu/1.63/www/Lec-notes/chap2_slow/2-5Stokes.pdf] for the pressure and velocity fields.

I am having trouble setting up the correct geometry for this problem. The way the problem is solved analytically is by using 2D spherical axisymmetric coordinates, involving r and \theta. The Laplacian for this case is different from the 2D cylindrical case (despite the latter being in terms of r and \theta) which is covered in the “Flow past a cylinder” example discussed in the tutorial [https://fenicsproject.org/pub/tutorial/pdf/fenics-tutorial-vol1.pdf].

Any help in getting started with the geometry setup is greatly appreciated.

Thank You

Have you looked at Solving PDEs in Python - <br> The FEniCS Tutorial Volume I ?
Test problem 2 treats flow around a cylinder.

Thank you, yes, I have looked at the problem. I am interested in flow around a sphere in 2D, and not flow around a cylinder. In my mind, the two need to be handled differently, because their Laplacians are different even though both use r and \theta coordinates.

I would be grateful for any help on handling 2D axisymmetric flow past a sphere.

Hi, here is an example of axisymmetric computation in solid mechanics which can help you adapting the Stokes flow problem to a spherical obstacle

You essentially need to redefine the gradient operator in polar coordinates and adapt the integration measure accordingly

1 Like

Thanks, Jeremy. Yes, I would have to redefine the gradient operator to handle spherical polar coordinates in 2D. I will give it a shot.