when i simulate a furnace,i find there are reverse flow in all outlet(pressure-outlet),why?who can tell me?and how to eliminate it during my case? second,the converge rate is very rapidly,when the iterate is 300,the varible is converged. thirdly,how i set the turb intensity and hydraulic diameter(the geometry is hexahedron)

bestregard

hanjun

(1). It depends on your furnace configuration, and initial flow field guess, assuming that the boundary conditions are all set properly. (2). You can extend the outlets (attach a pipe to it) to avoid the reverse flow problem. (3). As for the inlet condition, you have to read the user's guide. It is not difficult.

Hi,

I totally agree with what John Chien said in his messege. He did not discuss the convergence question though.

First of all, people usually do not complain when the solution converges nicely! You do however look at the definition of the convergence and the residuals used in the code. Fluent uses normailized (or is it scaled?) residuals as defaults. 3 orders drop in residuals is NOT a rule to judge convergence. I have infact seen quiet a few cases where the solution changed substantially after the 3 orders drop.

So I recommend: Read the section in the fluent user manual that discusses convergence. I usually try to get 4 orders drop (instead of the default 3). I also monitor other things, such as flow rate at the outlets (if its a multiple outlet problem) or and pressure at the inlet. You can do that using monitor/surface menu. I recommend turning this on after say 100 iterations, when the major initials fluctuations have been cleared.

thanks a lot