[ah_eshtiaghi]

i tried to solve a mixing flow
the first grid size is 1014815
the solution converged for this grid for turbulent intensity of 5-10%
I made two different grid by the same program
i changed the size very little
1:1051248
2:578124

the solution does not converge for each one.
what did i wrong?

[Ralf Schmidt]

Hi!

what is your turbulence model? are you using wall functions?

the grid size looks very high... what are the dimensions? What is your Re-number and what is the resulting yplus?

Ralf

[ah_eshtiaghi]

dimensions are
33x20x96 in

Re=1.7x10^5

it is k-e model
standard wall function

[Ralf Schmidt]

The values you gave are the total number of cells? If you have coarsened the grid so much (almost half its size), your cell quality might suffer. And this can causes problems with convergence. So take care, that the cells are not to much screwed!

For wall functions, your yplus value has to be in a certain range. Otherwise they are not valid (std wall funktions: 30 < yplus < 300). So, if your grid changes, your yplus changes as well. And the solver might have problems to calculate the wall function....

[ah_eshtiaghi]

i extracted the mesh from icem one more time.
in this time the solution is not converged and i dont know what is the problem. i solved the same problem last week and it was converged but it doesn't. i dont know what is the problem and i checked all the boundary conditions. i compared all summaries. both are same.
give me advice
thanks