What should we do if 2 out of 10 CFD models do not converge? We need to compare CFD results amongst all 10 cases but 2 of them not converged. If we change under-relaxation factors for 2 of these, will our comparison be valid? What do people normally do when encounter this problem?

I think the better way it's to change the underelaxation factor as you said...and try to find a result!
Next you can try to re-run the other 8 sim with the same underela...but just try one of those to see if its ifluence is important.
I hope for you that's not important!

I dont know if that's the right way to do it but that's how i did it. The problem was, everytime the urf's were changed back to a different value, the result also changes. So, which urf should we use, reliably? Am I the only one experiencing this kind of nightmare?

The solution should be independant of the route it takes to get there, the level of under relaxation should not matter, in the final solution.

You may need to look at your grid or boundary conditions.

allan

that is new... others?

This is not new, this is how it is. If it were not the case then how would you know what solution was the correct one for fdifferent values of relaxation. If your getting different solutions fo different values of relaxation then it could be that in some cases your relaxation is so tight that the solution is not changing from iteration to iteration. If this is the case you need to look at your grid.

allan

That's normal if you check your solution after a fixed number of iterations, because with lower urfs it will take more iterations to converge

It is not advisable to become too cocky if you are asking for help.

If this is new to you I recommend you get yourself a text book and brush up on your knowledge.