[Weiqiang Liu]

Hi all,

This thread is a very general thread. I've done combustion simulation with fluent for a very long time. Both gas phase and surface mechanism are included in the case, which makes it a complicated chemistry case containing normally 70 reactions.

My always problem is convergence. The temperature contours and species profiles are very close to literature results. However, what I always worry about is convergence issue.

Firstly, the residuals are never below 1e-6 like those author claimed in the paper. Besides, the residuals always oscillating with small magnitude.

Secondly, I tried to check flux report to determine convergence. I found mass balance is very easy to reached. However, energy balance is just like the residuals oscillating and varies in a relatively large magnitude.

To improve convergence performance, I tried to calculate a cold flow first and then includes chemical reaction. I tried to modify URFs. I also tried different chemistry solvers like stiff chemistry solver, CHEMKIN CFD solver, non explicit source. Or I tried velocity pressure coupling methods.

Unfortunately, convergence problem still exists.

Can anybody give me some suggestions or hints on how to improve convergence issues?

Please help!

Weiqiang