|April 3, 2010, 17:12||
Can we solve directly Stokes equations?
Join Date: Jan 2010
Posts: 2Rep Power: 0
Since the steady (2D) Stokes equations are linear, I have formed a matrix for both U, V and P and solved directly. The staggered grid was used. To avoid the case of singular matrix, pressure is fixed at a given point by add one more row(with only one non-zero element) to the matrix(so the matrix is non-square but still can solve)
I used this method to solve for a simple cavity problem with Utop = Uo.
When Uo = constant, I can get a good result.
But if Uo is changed along the top edge (Uo = Uo(x)) the result looks terrible!
Could you please give me any comments on the method I have used?
If my method is acceptable, please give me some suggestions to deal with the case of Uo = Uo(x).
Thank you very much.
Last edited by [thelight; April 3, 2010 at 17:33.
|Thread||Thread Starter||Forum||Replies||Last Post|
|Navier Stokes equations in rotation frame..?||vinayender||Main CFD Forum||2||December 1, 2009 01:12|
|the Stokes and Brinkman equations||yang||Main CFD Forum||4||January 18, 2006 07:20|
|Stokes algorithm?||Rui||Main CFD Forum||1||March 17, 2003 14:14|
|Computational complexity of Navier Stokes equations||Marco Ellero||Main CFD Forum||5||May 5, 1999 21:07|
|pouring a liquid||Ray Roberts||Main CFD Forum||6||February 5, 1999 15:07|