|February 26, 2007, 06:51||
Re: nonlinear convection-diffusion
Personally, I'd say that the simplest thing is to start from an initial guess (even 0) and then compute with one value and one unknown, for example
x * x => x_prev * x
where x_prev is the solution of previous iteration (or initial guess if it's first iteration) and x is the value that you actually compute.
If things goes well, you should converge the the real solution. When the x_prev is same that computed x, you converged.
Another solution is to make a linearisation by Taylor expansion, and then use the previous iteration as working point, from one iteration to anoter, as said above.
Hope it helps
|Thread||Thread Starter||Forum||Replies||Last Post|
|Moving mesh||Niklas Wikstrom (Wikstrom)||OpenFOAM Running, Solving & CFD||122||June 15, 2014 06:20|
|Estimate convection and diffusion terms in transport equation||Lance||CFX||0||April 4, 2011 11:16|
|Convection, diffusion and reaction||Joao||CFX||0||October 25, 2008 20:19|
|diffusion effect opposite to convection||hasan karabay||CD-adapco||0||December 5, 2006 07:59|
|Diffusion / convection||I. Dotsikas||Main CFD Forum||2||April 21, 1999 16:39|