Hi,

I have simulated turbulent non-premixed methane/air combustion in a furnace using the standard ke turbulence model and the eddy dissipation combustion model for two-step methane/air mixture.

The solution converges well for 1st order solution with default under-relaxation factors. however the URF's have to be decreased to 0.6 for species and 0.8 for energy in order to get a converged second order solution.

Are these URF's too low to condsider the solution valid?

What effect do these low URF's have on the solution?

Thanks for your help,

Mark

The under relaxation factors have no effect on the solution.

The URF's don't have any influence on the final converged solution, but do influence how many iterations it will take to reach that converged solution. In my experience, anything less than 1 for the energy equation takes VERY long to converge. Make sure your solution if fully converged. Monitor certain variables with iteration, in your case, for example, species flow rate and/or temperature at the outlet of your domain is a good one to monitor. If you are satisfied that your solution is not changing very much with further iterations, then consider it converged. Have your residuals flatlined?

I know for the eddy dissipation with multi-species, some under-relaxation below the default is required just to even get the solution TO converge. But, it takes a long time. Try and increase energy, O.8 is far too low. It will take forever to fully converge.