I've just begun to work on a gas combustion furnace, a cylinder with a burner on one end (1 inlet for gas, another for oxygen in the center) and the outlet on the other end. I could not get a nice convergence using eddy break-up model, in cfx4.4,

Compare to the furnace body, the size of the burner inlets are really small, the grid has a big size defference. Is it the reason caused the convergence problem? How can I get a better grid quality (and convergence) with this situation?

Hi,

I'm not sure, if i got it right. You have large cylinder, with a burner on one side. The burner inlets are small. I guess, you have symmetry conditions and modelling just a part of the cylinder. From my experience: make sure you have around 4-8 cells acros inlet. Estimate where are the regoins of mixing, try to make the grid dense there. look also at the outlet. I've found helpfull to have finer grid (in radial direction) also near the outlet. Cut the domain into more blocks, so the grid has smooth change in shape and aspect ratio. try to make a grid, which is following a flow-field.

other tricks i've found usefull: try CCCT scheme -works good with combustion. try deferred correction, try false time stepping.

good luck.

matej