# inviscid 2d computation for M=0.3 flow past cylinder on fine mesh can not converge

 Register Blogs Members List Search Today's Posts Mark Forums Read

 March 17, 2011, 10:16 inviscid 2d computation for M=0.3 flow past cylinder on fine mesh can not converge #1 New Member   Zhu Xiaofeng Join Date: Sep 2009 Posts: 4 Rep Power: 15 I try to use very fine mesh to test my Euler solver, however I get the following surprising results. I wish someone can give me a explain. Both my solver and Fluent can get a correct result on a coarse mesh. The Mach contour shows symmetrical. wake is very weak. Both my solver and Fluent can not get a correct result on a fine mesh. The Mach contour shows unsymmetrical, and wake is strong and long. Fluent even can not get a converged result! The diameter of the cylinder is 2.0. The mesh is structured. The first layer mesh is 0.001 for fine mesh; and 0.02 for coarse mesh. Far field boundary is 40 times of the diameter away. And I am quite sure that my solver and Fluent is correctly setted.

 March 20, 2011, 02:27 #2 New Member   Zhu Xiaofeng Join Date: Sep 2009 Posts: 4 Rep Power: 15 Can't any one help me, on such a simple problem ...?

 March 20, 2011, 03:41 #3 Super Moderator   Praveen. C Join Date: Mar 2009 Location: Bangalore Posts: 342 Blog Entries: 6 Rep Power: 16 Describe the numerical scheme, post pictures of grid, solution. Then somebody might be able to help.

 March 20, 2011, 05:26 #4 Senior Member   Martin Hegedus Join Date: Feb 2011 Posts: 500 Rep Power: 18 The issue may be that a subsonic Euler solution has very little dissipation. When the solution starts up a high pressure compression zone will occur at the front and a low pressure expansion zone occurs at the back side. The solution will then oscillate between the front and the back. Eventually these oscillations may die down. Maybe. The solution methodology may keep it going. If you are just testing out your code by comparing it to Fluent you can run it at a Mach number high enough to get a shock. Maybe Mach 0.7. The shock will isolate the back from the front. But it still may take time to converge.